1. science
2. medicine
3. psychology and psychiatry

Appendix A: Sample Questionnaire
Participant Information Sheet (Quantitative phase)
Psychology
School of Human & Community Development
University of the Witwatersrand
Private Bag 3, WITS, 2050
Tel: (011) 717 4500 Fax: (011) 717 4559
Dear Student
Hi! We are currently conducting research on the interpersonal relatedness factors of personality. This
study forms part of a larger study on personality in South Africa. Part of this research requires your
responses on the attached questionnaire. It should take you approximately 1½ hours to complete the
questionnaire. Please ensure that you answer on the answer sheet. I understand that this is a
substantial investment of your time. However your response is valuable as it will contribute towards a
South African understanding of personality and will have an impact on research nationally and
internationally. I would therefore like to invite you to participate in this research.
Your responses will remain confidential and anonymity is guaranteed. At no time will I know who you are
since the questionnaire requires no identifying information. Completion and return of the questionnaire will
be considered to indicate permission for me to use your responses for the research project. Should you
choose not to participate, this will not be held against you in any way. As I am only interested in group
trends, and have no way of linking any individual’s identity to a particular questionnaire, I will not be able to
give you individual feedback. Feedback in the form of a one-page summary sheet will be available on
request approximately six months after completion of this questionnaire. If you have any further questions
or require feedback on the progress of the research, please feel free to contact me. My contact details
appear below my signature. The data from this study will be used in conjunction with data collected from
other South African samples to determine the value of the South African Personality Inventory.
Thank you for considering taking part in the research project. Please detach and keep this sheet.
__________________
Dr. S. Laher
011 717 4532 / [email protected]
DIRECTIONS
This booklet contains statements about your interests, preferences, everyday thoughts and behaviours
and how you feel about things. You need to respond to all statements in the booklet. The questionnaire
contains 364 statements. Please use an HB pencil to complete the questionnaire. Darken the circle that
corresponds to the number that best indicates your answer:
1. Strongly Disagree (1)
2. Disagree (2)
3. Somewhat Disagree / Somewhat Agree (3)
4. Agree (4)
5. Strongly Agree (5)
When the administrator asks you to start, answer each statement. Please keep the following things in
mind:
1. Read each statement and choose the answer that best describes you. There are no “right” or “wrong”
answers; just mark what is true for you.
2. If you want to change an answer, please erase the original one completely and darken the circle of
the number that best indicates your answer.
3. Please answer truthfully.
4. Complete the biographical information.
Please ask now if you have any questions.
Appendix B: Focus Group Participation
Participant Information Sheet (Qualitative phase)
Psychology
School of Human & Community Development
University of the Witwatersrand
Private Bag 3, WITS, 2050
Tel: (011) 717 4500 Fax: (011) 717 4559
Dear Student
Hello! I am currently completing my Master’s degree in Industrial Psychology at the University of the
Witwatersrand. As part of my course I need to complete a research project. I am conducting research into
personality theory and personality assessment in the South African context. Part of this research requires
your participation in a focus group that should not last more than 1 hour. I understand that this is a
substantial investment of your time. However your response is valuable as it will contribute towards a
South African understanding of personality and is perceived to have an impact on research nationally and
internationally. I would therefore like to invite you to participate in this research.
I will be conducting the focus group and transcribing your responses thereafter. However, your responses
will remain confidential, as pseudonyms will be used to refer to the participants eg. X, Y. My supervisor
and I are the only people who will view your responses. Should you choose not to participate, this will not
be held against you in any way. You may withdraw your response at any time. There are no risks or
benefits associated with participation in the focus group.The study may be written up for publication in a
journal or presented at a conference. For this reason raw data will be stored in a locked cupboard on
campus for a period of up to three years and then destroyed. Feedback will be available 6-9 months after
completion of the questionnaire in the form of a one to two page summarywhich will be available on
request from me via email and will be placed on the first year psychology notice board outside the
psychology office. This will be available on request. If you have any further questions or require feedback
on the progress of the research, please feel free to contact me. My contact details appear below my
signature.
Thank you for considering taking part in the research project. Please detach and keep this sheet.
______________________
Leah Branco e Silva
________________________
SumayaLaher
0824517119
011 717 4532
[email protected]
[email protected]
Informed Consent Form (Focus Group)
INFORMED CONSENT FORM (Focus Group)
(Please detach this section and return to the researcher)
I, _____________________________________ hereby agree to participate in Leah Branco e
Silva’s research entitled “Exploring the collective dimension of personality using the SAPI and
CPAI-2 in a sample of South African students”. I have read and understood what participation
entails as set out in the information sheet. I understand that:
o
Participation in this study is voluntary
o
No information that may identify me will be included in the research report, and my
responses will remain confidential
o
I may withdraw information from the study at any time
o
There are no risks or benefits associated with the study
o
I may choose not to answer any questions that I would prefer not to answer
o
The study may be written up for publication in a journal or presented at a conference;
therefore raw data will be stored safely in a locked cupboard on campus for a period of up
to three years and then destroyed.
______________________________
_______________________
Signature
Date
Informed Consent Form (Recording)
INFORMED CONSENT FORM (Recording)
(Please detach this section and return to the researcher)
I, _____________________________________ hereby agree to participate in a focus group for
Leah Branco e Silva’s research entitled “Exploring the collective dimension of personality using
the SAPI and CPAI-2 in a sample of South African students” which will be video recorded. I
understand that:
o
The tapes and transcriptions will not be seen or heard by any other person other than the
researcher and her supervisor.
o
No information that may identify me will be included in the research report, subsequent
presentations and/or publications.
o
My responses will remain confidential.
o
The study may be written up for publication in a journal or presented at a conference;
therefore all video recordings will be stored safely in a locked cupboard on campus for a
period of up to three years and then destroyed.
o
Direct quotes from individuals in the focus group may be used in the research report,
presentations and/or publications, but these will be referred to using pseudonyms, e.g. X,
Y etc.
______________________________
_______________________
Signature
Date
Appendix C
Focus Group Questions
Hello. I am Leah. We spoke on the phone. I would like to thank you for agreeing to
participate in my study. Before beginning with the focus group, I would like to assure you
that everything you say during this focus group will be kept confidential, and only my
supervisor and I will have access to the video tapes. The study may be written up for
publication in a journal or presented at a conference; therefore raw data will be stored safely
in a locked cupboard on campus for a period of up to three years and then destroyed.
Although anonymity cannot be obtained, confidentiality will be maintained by not disclosing
any information that is of a personal nature in the report. Assigning a pseudonym to your
information in the report, for example, Respondent X or Respondent Y, will maintain
confidentiality.
I would like to remind you that you maintain the right to withdraw from the study at any time
during the focus group. You also have the right to refrain from answering any question
should you wish to do so. There are no benefits or risks associated with participation in the
focus group. A feedback sheet in the form of a one to two page summary of the study and its
findings will be provided to you upon request. The feedback will be available approximately
6-9 months after the collection of the datain the form of a one to two page summary which
will be available on request from me via email and will be placed on the first year psychology
notice board outside the psychology office. My contact details are on this letter which I
would like you to read now. Following this, I would appreciate it if you could please
complete the attached consent forms. The first is consent for participation in the focus group
and the second is consent for the group to be recorded.
Thank you. If you are ready we can begin the focus group.
1. What is your understanding of personality?
2. What is your understanding of personality in the South African context?
3. Do you believe there is a difference between personality in South Africa and
personality in other countries around the world?
4. What was your experience of filling out the questionnaire?
5. Did you find the language problematic in the questionnaire you answered in the South
African context?
6. Do you think that the questionnaire will be problematic for particular cultural groups
in the South African context?
7. Were there any other issues relating to personality or the questionnaire that we have
not discussed as yet?
Thank you for your co-operation!
1
Chapter 1: Literature Review
1.1. Introduction
“All the world’s a stage, and all the men and women merely players”
(As You Like It, Act III, Scene 7).
‘Personality’ or ‘persona’, in Latin, traces back hundreds of years to its origin within the
theatre. It is here that Theophrastus first created character sketches- descriptions of a type of
person recognisable across time and place (Allport, 1961; Friedman &Schustack, 2006).
These characters were displayed with a theatrical mask implying an external imposition of an
archetypal character onto an individual. This mask thus depicted an expectation of the
individual to behave in a particular way throughout a performance as dictated by the
prototypical character (Schultz & Schultz, 2008). The theatrical mask is depicted in the quote
above from Shakespeares’s production As You Like It (Friedman &Schustack, 2006).
Social philosophers, consequently, proposed that one’s true self is thus merely comprised of
different masks; alluding to the idea that there may be no core self or personality (Schultz &
Schultz, 2008). In contrast, however, our daily interactions with individuals seem to stand in
opposition to this view. Our experiences with other human beings result in the ascription of
one’s behaviour to their personality- an innate aspect of each individual (Laher, 2010). We
describe those around us in terms of these innate qualities dictated by their personality, using
adjectives like shy, quiet and responsible to capture the essence of what makes each
individual who they are (Laher, 2010). Personality can thus be defined as “the dynamic
organisation within the individual of those psychophysical systems that determine his unique
adjustments to his environment” (Allport, 1937; Robbins, Judge, Odendaal&Roodt, 2009, p.
92). In addition, the concept of personality is used to explain the consistency in one’s
behaviour over time and the distinctiveness in behaviour between individuals (Weiten, 2007).
Personality psychology incorporates eight key positions; these positions outline and describe
the main approaches to personality in psychology. Personality as a function of the
unconscious is the first key position adopted in the understanding of personality. Freud
2
proposes that one’s unconscious desires, thoughts and feelings are the primary determinants
of their behaviour. The second perspective emphasises the individual as influenced by ego
forces that characterise the ‘self’ and provide one with their identity (Friedman &Schustack,
2006). One’s biological being is the third perspective that influences individual behaviour;
this is one’s genetic, physical and physiological make-up, as well as their temperament. The
fourth perspective dictates that one’s behaviour is shaped and conditioned by their
environment and the experiences that surround them. Culture is a very influential aspect in
this respect, as it determines how we respond to certain situations (Larsen & Buss, 2008;
Schultz & Schultz, 2008). Individual cognitive dimensions outline the fifth perspective in
relation to personality psychology. This perspective incorporates the way one perceives and
interprets the world around them. Different individuals hold different mental representations
of this world. The sixth perspective is the trait approach to psychology and incorporates one’s
individual inclinations and abilities, and encompasses their traits, skills and predispositions.
The seventh perspective includes one’s spiritual dimension which prompts them to ponder the
meaning of their existence. The final perspective that affects individual behaviour is their
pursuit of happiness and self-fulfilment, found in interactions with the individuals in one’s
environment. The eight key positions detailed above closely resemble the eight theoretical
approaches outlined in personality psychology literature, in which all theoretical approaches
are subsumed; namely psychodynamic, the lifespan theories, cognitive, social learning,
humanistic/existential, the behaviourist approach, the biological/behavioural genetics
theories, or the dispositional/trait approach (Friedman &Schustack, 2006; Larsen & Buss,
2008; Schultz & Schultz, 2008).
This study focuses on one’s traits, skills and predispositions outlined by the trait approach to
psychology (Larsen & Buss, 2008; Schultz & Schultz, 2008).This is the basis on which the
personality frameworks that will be considered and discussed in this study were developed;
namely the Five-Factor Model (FFM). McCrae and Costa (1995) state that personality- and
its subsequent traits- are universal and thus transcend time, place and circumstance. These
trait structures that constitute an individual’s personality describe consistent patterns of
feeling, cognition and behaving (Church, 2000; John, Robins &Pervin, 2008).
3
Since the advent of democracy in 1994 rapid transformation is continuously occurring in an
attempt to redress the social inequalities resulting from apartheid. Psychometric assessment
played a substantial role in the previous political dispensation of South Africa, such that these
social inequalities were perpetuated and maintained (Laher, 2010). Psychometric tests were
developed and standardised in line with the unfair distribution of resources within the country
in order to serve the White elite. Thus these psychometric assessments were used as a means
with which to justify this unequal distribution and exploitation of Black labour (Nzimande,
1995; as cited in Laher, 2010); hence determining admittance to economic and educational
opportunities (Sehlapelo& Terre Blanche, 1996). Psychological assessments thus possess a
stigma that has resulted in the distrust of such assessments in organisations by minority
groups. It is for this reason(and more) that many organisations in South Africa, such as
Investec; do not make use of psychological assessments. Psychologists in South Africa today
are thus aware of the importance of creating and utilising tests that are fair and unbiased
(Sehlapelo& Terre Blanche, 1996; Laher, 2010; Nel, 2008). Legislation has been developed
in order to regulate the equal distribution of such opportunities in South Africa amongst its
many cultures and language groups. Section 8 of the Employment Equity Act (Office of the
President, 1998) regulates who will gain access to economic and educational opportunities by
stating that psychological testing is prohibited unless the assessment being used is valid and
reliable, applied fairly to all employees and is not biased against any employee or group
(Sehlapelo& Terre Blanche, 1996; Laher, 2010).
Personality is believed to underlie every aspect of organisational behaviour, and is a key
factor in determining organisational effectiveness. Consequently, personality assessment is
considered to be central to four aspects of organisational effectiveness; namely selection,
recruitment and retaining of talent; motivation; leadership and components of strategy
(Schneider & Smith, 2004). Personality characteristics are essential in making decisions
within an organisational setting; therefore emphasising the importance of personality
assessments (Weiner & Greene, 2008). A large body of methodological research from the last
two decades provides considerable support for the use of carefully constructed personality
tests in employment decisions across a variety of settings; such as selection, performance,
development and retrenchment. However, the most relevant personality dimensions in
relation to job performance for a particular job are still under investigation. The
4
appropriateness of an assessment for occupational use cannot be considered separate from the
purpose, situation and population involved in a given context (Anastasi&Urbina, 2007).
Psychological tests, including personality assessments, are commonly employed within
organisations for individual counselling; as well as occupational decisions concerning
selection, assignment, managing and evaluation of personnel (Sehlapelo& Terre Blanche,
1996). This consequently determines who will obtain access to organisational opportunities
(Owen &Taljaard, 1996; as cited in Van der Merwe, 2002). Research has demonstrated a
relationship between job productivity and the validity of an instrument, such that effective
selection methods can contribute substantially to overall organisational productivity. This
further emphasises the need to rectify the issue of job placement for previously disadvantaged
minorities within South Africa through the use of valid assessment measures
(Anastasi&Urbina, 2007). Tests should be used to “increase opportunities for members of
designated minority groups [and] should not be confounded with the validity of the testing
process” (Anastasi&Urbina, 2007, p. 550).
Personality assessment is used within organisations as a means with which to predict
academic and job performance (Roodt&La Grange, 2001;Nel, 2008). It has been found that
personality traits influence achievement at work, as personality characteristics correlate with
job success (Kline, 1992b; as cited in Kline, 1993; Foxcroft&Roodt, 2009). All five factors of
the FFM have been found to predict job performance; yet the facets of Conscientiousness
have been found to be the highest predictors of job performance (Guion&Highhouse, 2006).
Extraversion and Openness are important dimensions to understand in order to provide
insight into the client’s occupational directions. Individual motivation interacts with
knowledge and skills in order to impact on job performance; whereas individual motivation
and attitudes have been found to determine job satisfaction and organisational commitment
(Guion&Highhouse, 2006). Personality tests that have no connection to the requirements of
the job are inappropriate and unethical for use in selection and appraisal. Personality
assessment is also used to obtain individual differences for training, development and reward
(Foxcroft&Roodt, 2009). Performance and development within organisations have been
found to be a result of motivational drives located within personality; as well as interpersonal
factors such as sociability, dominance and flexibility. Similarly, personality assessments are
used in order to assess how individuals will learn and what qualities can be taught through
training and development (Lanyon & Goodstein, 1982).
5
The majority of personality inventories utilised in South African organisations are imported,
such that the languages and cultures that these assessments are adapted too are significantly
different from those where the assessment was developed (Nel, 2008). The use of imported
inventories presents an array of difficulties for personality assessment within South Africa.
Meiring (2006) states that the use of mainly English questionnaires is problematic in a
multilingual and multicultural society like South Africa, such that it results in the absence of
construct equivalence. South Africa comprises of eleven official languages providing an
assortment of primary languages for possible test takers; yet imported personality inventories
are usually translated to only English or Afrikaans within South Africa. This further
emphasises the social inequality still rife in South Africa as a result of the previous political
dispensation (Nel, 2008).
De Bruin, De Bruin, Dercksen, Cilliers-Hartslief (2005) reveal that due to the many African
languages present in South African society, it is likely that the first language of many
participants will be an indigenous African language. However, there are very few personality
assessments available in indigenous African languages; thus emerges English language
proficiency issues for participants completing imported English inventories. Indigenous and
collective personality traits are suggested to be absent from etic measures of personality; it
can thus be said that the Cross-Cultural Personality Inventory – 2 (CPAI-2) andSouth African
Personality Inventory (SAPI) weredeveloped as a result of this concern. These psychometric
instruments were therefore developed using the etic-emic approach in order to incorporate
those constructs missing from existing structures. As a result, this study explored reliability,
validity and bias of the collective dimension of the SAPI and CPAI-2 in a sample of first-year
South African psychology students.
The literature review that follows will delve into a brief history of personality, outlining the
emergence of the FFM. A discussion of this model, as well as its universal applicability will
continue- this will include the applicability of this theory in other cultures. Personality in
terms of individualistic and collectivistic dimensions will be explored, followed by a
discussion of the etic-emic debate within personality assessment. An introduction of the
CPAI-2 and SAPI (the instruments under investigation in this study) will be given. Lastly
issues of reliability, validity and bias with regard to these assessments will be discussed.
6
1.2. The Trait Approach to Personality
According to McCrae and Costa (1996) trait psychology can be described as the “core of
personality”. Traits are said to determine habits, attitudes, skills, beliefs, and other
characteristic adaptations (McCrae, 2001). They thus express stable and enduring individual
differences in thoughts, feelings and behaviour (Church, 2000). Traits can therefore be
declared to be situation-dependent in that they exhibit a predisposition to certain behaviour
(Tellegen, 1991; as cited in McCrae & Costa, 2008b). Traits form the basis for cross-cultural
and evolutionary research on personality, and thus it is argued that without traits the study of
personality and the psychometric approach would not exist (Wiggins, 1997 as cited in
Church, 2000).
The trait approach to psychology emerged from the lexical approach. This approach involves
factor analysis in which a number of personality adjectives are loaded to determine which of
the important individual differences encode as single terms (Goldberg, 1990). Sir Francis
Galton was the first scientist to use the lexical approach; compiling a lexicon of
approximately 18 000 terms to discover all the terms related to personality traits (Goldberg,
1995). Thurstone developed the statistical technique factor analysis in order to create a
comprehensive theory of personality by minimising 60 common trait adjectives to 5
independent common factors (Goldberg, 1990). In 1936, Allport and Odbert used lexical
analysis to extract 17 953 words describing stable personality traits from English language
dictionaries. Following from this they reduced the adjectives to 4 504 words that were
believed to best describe observable personality traits (Allport&Odbert, 1936).
Cattell developed on Allport and Odbert’s work, using their 4504 descriptive adjectives as the
basis for the minimisation of these adjectives to 171 words. This was done by eliminating
synonyms and adding words obtained from psychological research (Cattell, 1957). Semantic
and empirical clustering techniques were then used to reduce this list to 35 major clusters of
personality traits. These variables were then subjected to factor analysis from which 16
factors emerged; these factors formed the basis of Cattell’s 16-Factor Personality
questionnaire (16PF) (Cattell, 1957). With the use of Cattell’s 16 factors, Tupes and
Christal(1961) discovered five recurring personality factors;namely Surgency,
7
Agreeableness, Dependability, Emotional Stability and Culture. These five factors represent
the FFM. Norman replicated this work and found support for these overarching five factors.
However, the discovery of the FFM was largely ignored and only resurfaced 20 years later
(Tupes&Christal, 1961).
Theorists such as Digman, Costa, McCrae and Goldberg renewed interest in research on the
FFM, resulting in the aforementioned five factors being widely accepted by personality
theorists in the 1980s (Laher, 2010). These five factors of personality form the FFM and
account for the majority of variations in human behaviour (Costa & McCrae, 1992). The
NEO-PI-R was developed in 1990 as an operationalisation of the FFM and is the most widely
accepted psychometric test in this regard (Costa & McCrae, 1992). Nonetheless problems
have been encountered recently in replicating all the factors of the FFM in cross-cultural
linguistic data (Paunonen, Zeidner, Engvik, Oosterveld&Maliphant, 2000; Laher, 2008;
McCrae, Terracciano&78 Members of the Personality Profiles of Cultures Project,2005).
This was the premise on which the CPAI-2 and the SAPI were developed thereafter.
The next section of the literature review will examine the FFM- providing descriptions of the
five factors as well as its cross-cultural applicability and its application, in South Africa in
particular. This will then follow on to a description of the SAPI and CPAI-2- the two
instruments under investigation in this study.
1.3. The Five-Factor Model
1.3.1. The Five Factors
The FFM provides a simple and effective means by which trait structures can be described.
This particular trait structure is believed to be exhaustive, such that the five dimensions of the
FFM can be used to characterise every individual personality universally (Costa & McCrae,
2008). This organisation of personality traits represents individual differences in thoughts,
feelings and behaviours (Church, 2000). The model is organised into five domains:
8
Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness
(Costa & McCrae, 1992).
Neuroticism or emotional stability can be described as a tendency to experience negative
emotions such as fear, sadness, embarrassment, guilt, anger, and disgust. It is the extent to
which individuals are calm and self-confident, rather than anxious and insecure (Costa &
McCrae, 1992; Dalton & Wilson, 2000; McCrae, Yik, Trapnell, Bond and Paulhus, 1998).
Extraversion or Positive Emotionality is a tendency toward sociability, assertiveness, activity
and talkativeness. It is the degree to which a person is sociable, leader-like and assertive
(which reflects a high score on Extraversion) as opposed to withdrawn, quiet and reserved.
Extroverts like stimulation and excitement, and are generally cheerful and optimistic (Costa
& McCrae, 1992; Dalton & Wilson, 2000; McCrae, Costa, Del Pilar, Rolland & Parker,
1998). Openness to Experience entails active imagination, aesthetic sensitivity, attentiveness
to inner feelings, a preference for variety, intellectual curiosity, and independence of
judgment. It is the degree to which an individual is imaginative and curios, rather than
concrete minded and narrow thinking (Costa & McCrae, 1992; Dalton & Wilson, 2000;
McCrae, et al., 1998). Agreeableness refers primarily to the degree to which a person displays
altruistic tendencies towards others. It is the extent to which a person reflects good natured,
warm and cooperative behaviour as opposed to behaviour that is irritable, inflexible,
unpleasant and disagreeable (Costa & McCrae, 1992; Dalton & Wilson, 2000; McCrae, et al.,
1998). Conscientiousness concerns self-control and skills involved in planning, organising
and performing tasks. The conscientious individual is seen as purposeful, strong-willed and
determined. It is the degree to which a person is persevering, responsible and organised
which can be seen through achievements; as opposed to behaviour that is lazy, irresponsible
and impulsive (Costa & McCrae, 1992; Dalton & Wilson, 2000; McCrae, et al., 1998).
1.3.2. Cross-Cultural Applicability of the FFM
Many studies have been conducted in an attempt to explore the universality of the FFM using
the NEO-PI-R, as well as to replicate the proposed five factors. McCrae and Costa (1997)
compared the American NEO-PI-R personality test with that of German, Portuguese,
Hebrew, Chinese, Korean and Japanese factor structure and found that they were similar; thus
9
providing evidence of this universal structure of traits. Translated versions of the NEO-PI
have also been found to replicate well in Dutch, German, Italian, Estonian, Finnish, Spanish,
Hebrew, Portuguese, Russian, Korean, Japanese, French and Filipino cultures (Church
&Lonner, 1998). This replication has also been extended to unrelated languages, such as
Sino-Tibetan, Hamito-Semitic, Uralic, and Malayo-Polynesian (McCrae, et al., 1998). The
FFM and the NEO-PI-R have been translated into many European languages, but very few
African languages. Subsequently research has been performed in Africa in order to determine
whether the Western inspired FFM can be reproduced in a non-Western context. Yet
problems seem to arise for African samples as tests are not taken by participants in their
native languages, therefore affecting responses to test items (see Bedell, van Eeden& van
Staden, 1999; Cohen &Swerdlik, 2005; Foxcroft, 2004; Van de Vijver&Rothmann, 2004).
McCrae, Terraccianoand 78 Members of the Personality Profiles of Cultures Project(2005)
performed a study across fifty cultures in order to test whether the FFM is universal and is in
fact replicable in a non-Western sample. The sample included African and Arabic college
students and the 3rd person version of the NEO-PI-R was used. Factor analyses within the
different cultures indicated that the FFM appeared valid across all fifty cultures; however this
was not true for the five African cultures (McCrae et al., 2005). Allik and McCrae (2004)
conducted an international study of 36 cultures, including Black and White South Africans, in
order to test the same hypothesis. This study established that Black South Africans, as well as
Africans and Europeans, score lower on Extraversion and Openness to Experience and higher
in Agreeableness (Allik& McCrae, 2004). McCrae et al. (2005) performed another study that
explored the universality of the FFM in a sample of five African countries (Botswana,
Nigeria, Burkina Faso, Ethiopia, Uganda) and found that factor solutions in these cultures are
not perfect replications of the American normative structure of the FFM. The low
congruencies found in the above study were attributed to African personality structures that
differ from the FFM in a Western sample and/or to the problems that arise in using an
inappropriate Western questionnaire in this cultural context (McCrae et al., 2005).
10
1.3.3. Five or More Factors?
There has been much debate amongst researchers as to whether the five factors as proposed
by the FFM are the only factors describing personality, or whether more core factors do in
fact exist. McCrae and Allik (2002) reviewed a number of studies spanning 40 cultures and 5
continents; namely Indo-European, Altaic, Uralic, Hamito-Semitic, Malayo-Polenesian,
Dravidian, Austro-Asiatic, Sino-Tibetan, Japanese, Korean and Bantu languages. These
authors thus provide evidence that suggests that although the FFM does generalise adequately
across these cultures, there are specific variations between these said cultures. To name a few
of these variations: Church (2000) found evidence of a “Big Seven”-including Positive and
Negative valence,Digman and Takemoto-Chock (1981) found evidence of a sixth Culture
factor; Peabody and Goldberg (1989, as cited in Laher, 2010) discovered a Values factor, and
Krug and Johns (1986 as cited in Laher, 2010) suggested a sixth factor Intelligence. This has
prompted research into the proposed five factors by Cheung and Leung (1998) who found
evidence for a lack of the Openness factor in Chinese culture; while Ashton and Lee (2005)
found evidence to include a sixth factor named Honesty-Humility.
This was further qualified by African studies which suggest that five factors may be found,
however the essence of these factors may be different to the FFM (Laher, 2010). For example
in a sample of South African white-collar workers only three of the five factors were found;
namely Extraversion, Conscientiousness and Agreeableness (Matsimbi, 1997). In a study
conducted by Piedmont, Bain, McCrae and Costa (2002), the NEO-PI-R was translated into
the Zimbabwean language of Shona. The translation did yield the five-factor structure, but
more so for Extraversion and Agreeableness than the other three factors. Teferi’s (2004)
study aimed to explore the applicability of the NEO-PI-R in the Eritrean context by
translating the instrument into Tigrigna. Factor analysis was used and only Neuroticism,
Agreeableness and Conscientiousness could be extracted from the five-factor structure.
1.3.4. The FFM in South Africa
Studies using the NEO-PI-R in South Africa have also revealed differences in the five-factor
structure for race; such that a fit emerged for White participants but a lack of fit emerged for
11
Black participants. Taylor (2000) found that Openness to Experience could not be extracted
for the Black group, while in contrast all five factors could be found for the White group.
Another South African study administered the NEO-PI-R to 408 university students of
different race groups. Factor analysis with Varimax rotation revealed a personality structure
that was the same for all age groups; however the scores for some of the domains and facets
differed. Black students scored lower than White and Indian students in Openness to
Experience, whereas White students scored the highest on Extraversion and Agreeableness
(Heuchert, Parker, Stumpf&Myburgh, 2000).
A study performed by Heaven, Connors and Stones (1994) presented 230 South Africans
(predominantly Black) and 186 Australian students with a list of 112 adjectives that described
the FFM. A five-factor solution was found for the Australian sample but not for the South
African sample. Reasons for this result may include difficulties in understanding and
interpreting the instrument for participants whose primary language is not English, response
sets may be stronger for South Africans or the FFM does not adequately represent the
personality structure of Black South Africans (Piedmont et al., 2002). This reasoning is the
foundation for the development of the SAPI (Nel, 2008). McCrae et al. (2005) found a poorer
replication of the five factor structure for African samples than for Americans in Botswana,
Nigeria, Burkina Faso, Ethiopia and Uganda. Botswana, Nigeria, Ethiopia, Uganda and
Morocco revealed poorer data quality scores and internal consistency than the Western
nations in the study, such as Burkina Faso- who were able to use a French version of the
NEO-PI-R. These differences are attributed to the lack of translations for African cultures
compared to those of Western cultures (McCrae et al., 2005). Heaven and Pretorius (1998)
translated the NEO-PI-R into Afrikaans and Sotho; the normative five factor structure was
found for the Afrikaans-speaking group, though a different pattern emerged for the Sothospeaking group withthe clearest factor being emotional stability.
1.3.5. Problems with the FFM
All these studies mentioned above have raised a number of issues with regard to the
replicability of the FFM structure and the use of the NEO-PI-R in non-Western samples.
These problems include language, race and culture; particularly in a multicultural and
12
multilingual South Africa nation.The dimensions of the FFM are derived from descriptive
adjectives in the English language. It thus seems that language plays a substantial role in
whether or not the FFM is replicable cross-culturally (Ashton et al., 2004), as language has
been found to be a powerful moderator of test performance (Foxcroft, 1997; Heaven &
Pretorius, 1998).There has been some evidence for the universality of the FFM; however
these factors may manifest differently in different cultures (Yik& Bond, 1993). “McCrae
himself acknowledged that not all factors of the FFM might be equally important in every
culture” (McCrae, 2002; as cited in Cheung, Cheung, Zhang, Leung, Leong & Huiyeh2008,
p. 84). Thus its applicability across cultures, and particularly in Asian and African cultures, is
dubious (Church, 2000; Laher, 2010;Heaven & Pretorius, 1998; McCrae et al., 2005). This is
thus the premise on which the development of the SAPI and CPAI-2 were prompted.
1.4. Individualism vs Collectivism
Different cultures have been described as possessing individualistic and/or collectivistic
traits; thus determining whether the individual or society is of the utmost importance. It has
been suggested that the West (Western European, North American, Australian) is
characterised by a focus on individualism; while the non-West (Asian, South American,
African) is characterised by collectivistic traits (Hermans&Kempen, 1998; as cited inLaher,
2010; Green, Deschamps&Paez, 2005). Individualistic groups have been described as the ‘I’
and ‘me’ culture; while collectivistic cultures have been seen as the ‘us’ and ‘we’ groups
(Triandis, 1995). Individualism characterises the self as egocentric, with a focus on
independence, autonomy and personal goals. Collectivism, in contrast, characterises the self
as sociocentric, with a focus on interdependence, unity, loyalty and the achievement of
collective aims(sometimes at the expense of personal aims). Collectivistic cultures evaluate
themselves in terms of others, their interpersonal roles and the quality of their relationships;
and so they behave according to group norms rather than personal attributes (Green et al.,
2005). Euro-Americans are thus believed to maintain their independence through expressions
of their individuality and inner attributes rather than group norms; while non-Western
cultures, such as Asians, seek to maintain harmonious relationships assimilating in to the
collective group (Markus &Kitayama, 1998;Triandis, 1989).
13
The collective construction of personality refers to relationality, while individualistic
constructions view and describe groups in terms of individual identities submerged within a
social identity (Ellemers, Spears &Doosje, 2002;Markus &Kitayama, 1998). It has been
argued that personality traits are a Western phenomenon indicative of the ideal of
individualism, whereas these distinctive enduring traits do not characterise non-Western
collectivist personalities. These personality characteristics are instead determined by
interpersonal and momentary situations (Cross & Markus, 1999; Laher, 2010). This
distinction was made between North Americans and the Chinese in numerous studies utilising
the Twenty Statements Test. North Americans were found to describe themselves in terms of
trait terms, such as smart, open-minded and dependable, whereas the Chinese describe
themselves in terms of interpersonal roles such as friend or daughter (Ip& Bond, 1995; as
cited in Nel, 2008; Rhee, Uleman, Lee & Roman, 1995). These findings correlate with those
found in a South African sample. The English language group tends to describe others in
conceptualised terms such as patient, caring or respectful; whereas the Siswati language
group uses more socially descriptive terms like she is patient with her husband or she always
greets elderly people (Nel, 2008).
Increasingly it has come to the forefront that Western structures of personality are ill
equipped to encompass the indigenous traits of non-Western cultures. One such structure that
has been debated is the FFM of personality and has been discussed above (Costa & McCrae,
2008). The SAPI and the CPAI-2 are two personality assessments developed by
psychologists to satisfy the need for indigenous personality constructs and measures, instead
of adapting Western developed personality assessments. Studies using the CPAI-2 reveal the
existence of the five factors, as outlined by the FFM. However a sixth factor has also
emerged, namely Interpersonal Relatedness (Cheung, et al., 2003). This emergent factor has
left Cheung, et al. (2003) to believe that there is a gap in personality theory for indigenous
collective dimensions of personality that are not incorporated in Western personality theory.
This collective dimension describes an indigenous cultural context that involves the
achievement of equilibrium through the qualities of humanity and social propriety (Cheung et
al., 2008).
14
The SAPI is in the process of being developed in order to overcome current problems facing
personality assessment in South Africa. Ubuntu is an indigenous concept that originated in
South Africa to describe a ‘South African’ way of thinking (Nel, 2008). This concept
encompasses an interpersonal element and collective spirit of South Africans. The term
Ubuntu (humaneness) possesses a collectivistic orientation and describes a person as a person
through others (Louw, 2002; Nel, 2008). Ubuntu describes a respect for others,
belongingness, tolerance, compassion, obedience, loyalty, generosity, honesty,
trustworthiness and warmth.The concept of social relatedness is a key concept underlining
that of Ubuntu; as well as social analysis, social peace and social kindness in a collective
environment (Nolte-Schamm, 2006). Ubuntu as described by Louw (2001) is not
conclusively a collectivistic dimension of personality, but rather it describes the functioning
of an individual within a community, thus emphasising the community. “Ubuntu stems from
the philosophy that community strength grows from community support and that dignity and
identity are achieved through mutualism, empathy, generosity and community commitment,”
(Swanson, 2007; as cited in Laher, 2010, p. 48).
The SAPI does not make use of the term ‘Ubuntu’, as it incorporates many clusters that are
not unique to the eleven languages of South Africa. It was been suggested by Nel (2008) that
Bantu languages possess more collectivistic traits; while English and Afrikaans languages
possess more individualistic traits. This was also found in a study by Eaton and Louw (2002),
who discovered that African-language speakers, as compared to English-language speakers,
used more interdependent personality descriptive terms characteristic of the collectivist
dimension. While this may be true, to describe African language groups as collectivistic and
English and Afrikaans language groups as individualistic would be an inadequate
generalisation. South Africans may function in collectivistic and/or individualistic
environments and thus they adopt the necessary qualities of each environment depending on
the situation (Nel, 2008). However, Ubuntu is common in all language groups in South
Africa, though not exclusive to them. Ubuntu is measured in the SAPI through the clusters of
Facilitating, Integrity, Relationship Harmony and Soft-Heartedness. All South Africans have
been shown to possess a strong inclination to be sensitive towards others and to build and
maintain a community; thus revealing a collective dimension (Nel, 2008).
15
Nel (2008) proposed some correspondence between the clusters of the SAPI and factors of
the CPAI-2. An arbitrary relationship was found between Relationship Orientation and
Harmony facets of the CPAI-2 and the cluster of Relationship Harmony of the SAPI. The
definitions of the sub-clusters Interpersonal Relatedness and Approachability of Relationship
Harmony correspond somewhat to the definitions of the Relationship Orientation and
Harmony facets of the Interpersonal Relatedness factor of the CPAI-2. Approachability also
incorporates many facets that indicate socially responsive and sensitive behaviour
characterised by the Social Sensitivity facet of Interpersonal Relatedness. The Egoism subcluster of Soft-Heartedness of the SAPI contains the facet Generous, which encompasses
thrifty behaviour characterised by the facet Thrift vs Extravagance of the CPAI-2 (Nel,
2008). This reveals a minimal correspondence between the collective dimensions of the SAPI
and CPAI-2, thus revealing some relation between collective dimensions of personality for
South African and Chinese cultures. However, this also reveals that these collective
dimensions measure indigenous constructs of South African and Chinese personality that
may be distinct to these cultures. This study thus aimed to explore reliability,validity and bias
of the collective dimension of personality for the SAPI and the CPAI-2 that these assessments
propose to measure.
1.5. The Etic-Emic Debate
Etic constructs are universal and allow for comparisons between assessments of different
cultures. Emic constructs, on the other hand, describe constructs that are culturally specific
and thus go beyond the scope of the etic approach (Laher, 2010). The etic approach assumes
that most individuals around the world can be described using universal personality traits
(Cheung et al., 2003). It is for this reason that there is a dominance of imposed-etics, which
adapt and translate foreign English language tests for use in all cultures (Cheung et al., 2001).
Yik and Bond (1993) explicate that the inadequacy in using an imported-etic approach lies in
the omission of indigenous culture-specific personality traits found in the emic approach.
Thus although this may be a convenient means with which to make cross-cultural
comparisons “it does so at the possible cost of the validity of the test and does not necessarily
say anything meaningful about an individual in his or her world” (Cheung, Leung, Fan, Song,
Zhang, & Zhang, 1996; Laher, 2010, p. 67).
16
The use of the etic approach in South Africa would prove to be inadequate and problematic as
imported-etic measures are not representative of all eleven languages in South Africa (Nel,
2008). These problems include language proficiency and understanding of the items. An
additional factor that complicates the use of etic measures is cultural factors. It is believed
that personality structure may be universal; however it is expressed in different ways in
different cultures (Meiring, 2006). Thus the use of the indigenous or emic approach would be
suitable for the rich cultural context of South Africa when developing a personality
assessment (Nel, 2008). The emic approach obtains its value through its reflection of the
relation between human experiences and its definition of the self in a social and interpersonal
context (Ho, Peng, Lai & Chan, 2001). Swartz and Davies (1997) emphasise the use of the
emic approach in South Africa in order to deal more effectively with transformation in South
Africa and to improve diversity management in organisations.
Current research makes use of an etic-emic approach in order to capture these indigenous
cultural differences (Nel, 2008). This approach is valuable as the emic approach allows for
the indigenous personality-descriptive factors to be identified; while the etic approach allows
for comparisons between these indigenous dimensions and other cultures (Saucier &
Goldberg, 2001). This study thus made use of two psychometric assessments developed to
obtain an indigenous personality structure and assessment using an etic-emic approach,
namely the SAPI and CPAI-2. The personality structures derived in these instruments were
developed from everyday conceptions of personality, as well as through the review of
literature, surveys and descriptions of oneself and others in the South African and Chinese
context (Nel, 2008). Both the SAPI and CPAI-2 were developed based on a need for an emic
measurement of personality in a South African and Asian context, respectively. Although the
CPAI-2 will be an etic instrument in a South African sample, the use of both these
instruments will help explore their utility of the collective dimension of personality in South
Africa and will contribute to the understanding of personality and its development in the
South African context. The construct validity, internal consistency reliability and construct
bias of the SAPI was explored in order to validate the collective dimension of the SAPI in a
South African context. The construct validity, internal consistency reliability and construct
bias of the CPAI-2 was explored in order to determine the cross-cultural applicability of the
collective dimension in a South African sample (see Paunonen and Ashton, 1998).
17
1.6. The South African Personality Inventory
The South African Personality Inventory (SAPI) is a project established for the purpose of
overcoming the many issues currently experienced with regards to personality assessment in
South Africa. This project can be divided into two stages; namely qualitative and quantitative
(Nel, 2008). The qualitative stage is conceptual and is aimed at developing a set of common
and culture-specific personality traits in all eleven South African language groups for the
purpose of establishing traits that are relevant, accurate and authentic in all language groups.
The second stage aims at creating instruments based on the taxonomy derived in the previous
stage in order to develop a comprehensive inventory for personality assessment for all
language and cultural groups within South Africa (Nel, 2008).
The SAPI project adopted a type of lexical approach, instead of conceptualisations from the
FFM, in the acquisition of personality structures for all South African language groups
through everyday conceptualisations of personality (Nel, 2008). The personality-descriptive
terms were derived through semi-structured interviews in the participant’s native language. In
an attempt to combat issues regarding bias in previous personality assessments; participants
ranged from a variety of ages, education, linguistic and cultural groups, socioeconomic status
and urbanisation status (Nel, 2008). Participants were asked to describe “himself or herself,
another person opposite than him or her, a parent, a grandparent, a person from another
ethical group, a friend, a child (or a sibling), a neighbour, a teacher he or she liked, and a
teacher he or she did not like” (Nel, 2008, p. 56). These descriptions were provided by
participants in either behavioural or trait terms, creating a more accurate depiction of
personality structures within each language group. Thus allowing for a representation of the
similarities and differences in personality between all eleven language groups, and in turn
allowing for the development of a unified personality inventory that can be applied fairly to
all language groups within South Africa (Nel, 2008). The SAPI encompasses the universalspecific personality traits that were derived from the trait descriptions of all eleven language
groups (Nel, 2008).
More than 50 000 personality-descriptive terms were gathered from the interviews with
participants. These terms were interpreted and clustered using content analysis, and other
18
external resources, into 190 facets. These facets were then classified as common to all eleven
language groups, semi-common (found in 7-10 language groups), semi-specific (found in 2-6
language groups) and language-specific (unique to one language) (Nel, 2008). The 190 facets
were then further divided into 9 overall clusters with 31 sub-clusters (Nel, 2008). The 9
clusters include Extroversion, Soft-Heartedness, Conscientiousness, Emotional Stability,
Intellect, Openness, Integrity, Relationship Harmony and Facilitating (Nel, 2008).
The first six clusters (Extroversion, Soft-Heartedness, Conscientiousness, Emotional
Stability, Intellect and Openness) coincide with existing structures of personality and are thus
recognisable; however the last three structures (Integrity, Relationship Harmony and
Facilitating) seem to be somewhat unique to the SAPI (Nel, 2008). The quantitative phase of
this project is currently taking place and involves administering 2500 items to 4 language
groups in South Africa. Results, however, have not been published (Meiring, 2010). The aim
of this study is to explore the collective dimension of the SAPI which include the socialrelational items; Soft-Heartedness, Integrity, Relationship Harmony and Facilitating. These
are representative of the ‘Ubuntu’ concept and encompass the indigenous collectivistic traits
specifically designed for a South African sample (Nel, 2008).
The cluster Soft-Heartedness is comprised of the sub-clusters Amiability, Egoism,
Gratefulness, Hostility, Empathy and Active Support. Soft-Heartedness is defined as a
concern for the welfare of others, with a low concern for one’s own welfare and interests. A
person high in this cluster can be classified as compassionate and sensitive towards the
feelings of others and embraces the concept of community (Nel, 2008). The cluster Integrity
is compiled from the sub-clusters Integrity and Fairness. This cluster can be defined as a
human being’s moral consciousness. The characteristics of this consciousness include loyalty,
honesty and dependability (Nel, 2008). The cluster Relationship Harmony is compiled of
Approachability, Conflict-seeking, Interpersonal Relatedness and Meddlesome. Relationship
Harmony can be described as the belief that one should keep good relationships with others
through understanding, tolerance and keeping the peace (Nel, 2008). The final cluster
Facilitating is made up of the sub-clusters Guidance and Encouraging Others. This cluster
involves the guidance of others through example, advice and experience; as well as to
encourage others through one’s own behaviour(Nel, 2008).
19
This study intended to explore reliability, validity and bias of this personality structure by
exploring these social-relational dimensions- Soft-Heartedness, Integrity, Relationship
harmony and Facilitating- of the SAPI in a sample of first-year psychology students at the
University of the Witwatersrand.
1.7. The Cross-cultural Personality Assessment Inventory-2
Cheung et al. (1996) developed an instrument called the CPAI based on the emic-etic
approach that was designed to assess Chinese Personality from an indigenous perspective
(Cheung et al., 1996; Cheung et al., 2001). The personality constructs included in the CPAI
were derived from personality adjectives or personal descriptions of personality from
everyday life- from Chinese novels and proverbs, reviews of Chinese literature and surveys
of personal descriptions (Cheung et al., 1996). The main constructs of the CPAI are
Dependability, Social Potency, Individualism and Interpersonal Relatedness (Cheung et al.,
1996). The structure of the CPAI showed good correspondence with that of the FFM (Costa&
McCrae, 1992); however the construct of Openness to Experience was found to have the least
representation in Chinese culture (Cheung et al., 2008).
Interpersonal Relatedness is the sixth factor to come out of the CPAI and represents an
indigenous Chinese personality structure in a culturally meaningful way (Cheung et al.,
2008); thus it can be construed as non-Western traits (Cheung et al., 2003). This factor
emphasises the importance of social values within the Chinese collectivist culture.
Interpersonal Relatedness can be explained in terms of Confucian philosophy; the qualities of
humanity and social propriety manifesting as “inner sageliness and outer kingliness” (Cheung
et al., 2008, p. 92). Equilibrium is reached through self-discipline and personal cultivation
(inner sageliness), as well as loving people by behaving according to propriety (outer
kingliness) (Cheung et al., 2008). Interpersonal Relatedness showed little correspondence
with the FFM, thus revealing the cultural specificity of this factor to the Chinese culture.
However, a study by Lin and Church (2004) found the Interpersonal Relatedness factor in
Chinese American and European American groups, thus making this factor less specific to
the Chinese culture.
20
The CPAI was translated into English using a technique called back translation outlined by
Brislin (1970; as cited in Laher, 2010). The English version of the CPAI scales were then
administered to a multi-ethnic sample of Hawaiian undergraduate students (Gan, 1998; as
cited in Laher, 2010). The CPAI was then administered in two studies to determine its
applicability- this included 675 Singaporean Chinese and 144 Caucasian American students
at a Midwestern university, where similar factor structures were obtained in both samples.
However, it was found that the Singaporean sample was more like the Chinese normative
structure than the American sample (Cheung et al., 2003). The CPAI was administered with
the NEO-FFI and it was revealed that the Interpersonal Relatedness factor of the CPAI was
not covered by the NEO-FFI and the Openness factor of the NEO-FFI was not covered by the
CPAI (Cheung et al., 2003). This resulted in the development of the CPAI-2. Cheung et al.
(2003) revised the CPAI, changing the Individualism factor to Accommodation; this version
is known as the CPAI-2. The revision took place in order to increase the clinical applicability
of the scale, and aesthetics was added in order to represent the Openness scales. Thus it can
be said that this assessment has demonstrated factors similar to the FFM, but it includes
indigenous collective dimensions specific to the Chinese culture. Therefore it would prove to
be interesting to test the CPAI-2 within a collectivist South African sample.
The CPAI-2 has 28 personality scales, which factor into 4 factors- Social
Potency/Expansiveness (8 scales), Dependability (9 scales), Accommodation (5 scales) and
Interpersonal Relatedness (6 scales) (Cheung, 2004). This study will only be exploring the
collective dimension of the assessment, namely the Interpersonal Relatedness scales. This
factor comprises of six scales- Traditionalismvs Modernity, Ren Qing (Relationship
Orientation), Social Sensitivity, Discipline, Harmony and Thrift vs Extravagance (Cheung,
2004; Cheung et al., 2008). Interpersonal Relatedness measures various aspects of
“interdependence and reciprocity in traditional social relationships” (Cheung et al., 2001, p.
425).
Traditionalism vs Modernity measures the degree of one’s individual modernisation; thus it
explores one’s attitude toward Chinese beliefs and values in areas such as family
relationships, filial piety, rituals and chastity. High scores indicate individuals who are
obedient, conservative, protect old customs and oppose new ideas; while low scores indicate
21
individuals who challenge traditional concepts and superstition, and advocate democracy and
personal freedom (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001). Ren Qing or
Relationship Orientation measures the degree to which individuals adhere to cultural norms.
High scores indicate individuals who actively strengthen their relationships, change their
principles for the needs of others and help others strengthen their social ties; and low scores
reveal individuals who are disinterested in social customs, and avoid interactions and the
exchange of favours with others (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001).
Social Sensitivity measures the extent to which individuals are sensitive to others- including
qualities such as empathy, understanding, willingness to listen and communicate, as well as
those who are easily approachable (Cheung, 2006; Cheung et al., 2001; Cheung et al., 2008).
Discipline, on the other hand, examines the extent to which individuals are rigid and
disciplined, rather than flexible and adaptable. Furthermore, this is the degree to which
individuals are stubborn and organised, who dislike unpredictable events and like convention;
or individuals who adapt to changes in the environment and the situation while discarding
formality (Cheung, 2006; Cheung et al., 2001). Harmony is the degree to which an individual
is content and possesses inner harmony and peace of mind. Thus these individuals may be
said to possess virtues such as being even-tempered and peaceful, instead of confrontational,
insensitive and open to conflict (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001).
The Thrift vs Extravagance factor measures one’s carefulness in spending. High scores
describe thrifty individuals who maintain a simple life and are therefore careful in their
expenditure of money and resources. Low scores describe extravagant individuals who spend
a lot of money and flaunt their wealth (Cheung, 2006; Cheung et al., 1996; Cheung et al.,
2001).
This study intended to explore reliability, validity and bias of this personality structure by
exploring the Interpersonal Relatedness dimensions- Traditionalismvs Modernity, Ren Qing
(Relationship Orientation), Social Sensitivity, Discipline, Harmony and Thrift vs
Extravagance- of the CPAI-2 in a sample of first-year psychology students at the University
of the Witwatersrand.
22
The two instruments described above- the SAPI and CPAI-2- are under investigation in this
study. This investigation included an examination of the reliability, validity and bias of these
instruments. These aspects are discussed in the sections to follow.
1.8. Examining the Reliability of the SAPI and CPAI-2 Collective
Scales
Murphy and Davidshofer (2005) define reliability as the consistency of test scores; such that
the same person that is re-examined on a test or an equivalent test at a later stage will produce
the same or similar scores (Anastasi&Urbina, 2007). Moerdyk (2009; as cited in Laher, 2010)
includes in his definition the degree to which test scores are free from random error.
Reliability can be categorised to include; test-retest reliability, parallel/alternate forms
reliability, internal consistency reliability and inter-rater reliability (Anastasi&Urbina, 2007;
Murphy &Davidshofer, 2005; Rust and Golombok, 1989).
A number of constraints arise when testing for test-retest reliability, parallel/alternate forms
reliability and inter-rater reliability as a result of the design of this study; therefore only
internal consistency reliability was considered. Split-half reliability describes the splitting of
a test into two equivalent halves, obtaining scores for each half and correlating them (Murphy
&Davidshofer, 2005). According to Moerdyk (2009; as cited in Laher, 2010) internal
consistency reliability can be defined as the mean of all split halves. This is usually done
using the method of alpha reliability coefficients. Thus, in this study this was done using
Cronbach’s alpha coefficients for each of the collective dimension scales of the SAPI and
CPAI-2.
23
1.9. Examining the Validity of the SAPI and CPAI-2 Collective
Scales
The validity of a test involves determining what that test measures and how well it does so.
There are different forms of validity; namely face validity, content validity, criterion validity
and construct validity (Anastasi&Urbina, 2007; Rust &Golombok, 1999). This study
examined the construct validity of the SAPI and CPAI-2. Construct validity of a test concerns
whether a test measures what it purports to measure. The goal of this validation was therefore
to establish if test scores provide a good measure of a specific construct, as construct validity
is concerned with the extent to which a particular instrument may be said to measure a
construct or trait (Murphy &Davidshofer, 2005). Construct validity has thus been said to be
the primary validation of trait-based psychometric assessments (Larsen & Buss, 2008;
Schultz & Schultz, 2008). These constructs or traits stem from a closely related network of
associated theory. This makes construct validity the prime validation for the trait approach to
personality, which is the premise on which the SAPI and CPAI-2 were developed. Therefore
it was the most suitable form of validation for this study (Kline, 1993; Murphy, 2003).
This study thus involved assessing whether the SAPI and CPAI-2 in fact measure the
collective social-relational and Interpersonal Relatedness dimensions respectively.
Exploratory factor analysis was developed as a means for identifying psychological traits
(Laher, 2010; Van de Vijver& Leung, 1997) and was accordingly used to measure the
construct validity of these two psychometric personality assessments. Factor analysis, in
terms of personality inventories, can be defined as a technique used to identify common traits
that underlie a large number of items included in a personality assessment (Laher, 2010).
1.10. Examining Bias in the SAPI and CPAI-2 Collective Scales
The system of Apartheid, instituted in 1948, resulted in economic and political privilege for
White South Africans only (Heuchert, et al., 2000). White workers experienced preference
over jobs in the workplace- using psychometric assessments as tools to ensure the superiority
of Whites in organisations. This was done by administering tests that were developed and
standardised on White South Africans to mostly illiterate, uneducated or poorly educated
24
Black South Africans. As a result mistrust for psychological testing resulted amongst the
Black population (Foxcroft&Roodt, 2009; Sehlapelo& Terre Blanche, 1996). Due to the
controversial role played by psychometric assessments in South Africa’s history of inequality
and conflict, it is becoming increasingly important to redress these issues in research (Laher,
2010).
The advent of democracy in 1994 allowed for the resultant social, political and economic
inequalities to be redressed. The Employment Equity Act (Office of the President, 1998) was
developed in order to address this discrimination in the workplace- andso it states that all
psychological instruments should be reliable, valid, unbiased and fair for all groups in South
Africa. As outlined in the above Act, psychological assessments such as the SAPI and CPAI2 must be unbiased for all groups in South Africa- this includes race groups, language groups
and gender. Thus because of South Africa’s great diversity it is imperative that such tests are
unbiased and equivalent across culture and language groups (Abrahams &Mauer, 1999;
Foxcroft, 1997; Sehlapelo&TerreBlanche, 1996; Van Eeden&Prinsloo, 1997).
When systematic errors in measurement or prediction of a test exist, this test is said to be
biased (Murphy &Davidshofer, 2005). Bias in a test therefore exists if the testing procedure is
unfair to a group of individuals who can be defined in some way (Rust &Golombok, 1999).
This can be done by looking at a test’s data to determine the extent to which the test contains
bias measures. Poortinga (1989; as cited in Laher), on the other hand, describes bias as the
presence of nuisance factors in cross-cultural measurement. Van de Vijver and Tanzer (1997)
reveal that bias occurs when there is a mismatch between score differences in indicators of a
construct and the underlining trait; this could be due to characteristics of the instrument in
cross-cultural comparisons. Three types of bias are said to exist; namely method bias,
construct bias and item bias (Van de Vijver& Leung, 2001). Method bias refers to all sources
of bias as a result of the method and procedure of the study; including mode of
administration, instrument differences, sample incomparability, and tester and interviewer
effects. Construct bias refers to the bias that occurs when the construct measured is different
across groups. Item bias is the generic term for all disturbances at item level; if participants
have the same view on an underlying construct but do not have the same item score (Van de
Vijver&Tanzer, 1997). Construct bias was assessed in this study. Construct bias was
25
examined in order to determine if there was a difference between groups (race, home
language and gender) on each scale for the SAPI and CPAI-2. Item bias and method bias
were not examined in this study.
1.10.1. Race and Personality
South Africa represents a hub of cultural diversity with many population groups in its
borders. The country’s history of Apartheid reveals the use of psychometric tests as a means
with which to ensure poor performance for Blacks by providing them with tests standardised
for the White population group. These low scores where then used as a justification for the
inferiority of the Black race group (Foxcroft&Roodt, 2009). For this reasonrace is essential to
take into consideration when investigating the cross-cultural applicability of psychometric
instruments. A study done by Abrahams (1996) in South Africa revealed the possible
moderating effect of race on personality profiles as significant differences were found in the
means, reliability coefficients, factor structures and the way the items were answered for
different race groups- particularly Blacks and Whites. Other research using the FFM has
revealed that Black South Africans scored lower on Extraversion and Openness to
Experience, and higher on Agreeableness- revealing that these differences in personality may
be due to differences in expressions of personality in individualistic and collectivistic cultures
(Allik& McCrae, 2004). Piedmont et al., (2002) extend on this by relating these differences to
response styles and biases between Western and non-Western cultures. They argue that
collectivist and more non-Western cultures tend to avoid extreme answers as opposed to
Western or collectivist cultures. Some of these differences in the representation of the FFM in
different race groups have been cited in the FFM section above.
1.10.2. Home Language and Personality
Population group and language have been found to be influential moderators of test
performance in terms of personality (Laher, 2010; Meiring, Van de Vijver,
Rothmann&Barrick, 2005; Nel, 2008). Research on the FFM has explored the universality of
the model and found disagreement in its applicability across cultures. Some researchers have
discovered issues in terms of language as it relates to personality, as some personality
26
descriptive adjectives in English and the consequent constructs have not been found in all
other cultures (Allik& McCrae, 2004; Ashton & Lee, 2005). Language in South Africa’s
history was used as a political tool to achieve superiority; socially and organisationally. Thus
language is a very important consideration when exploring the utility of any psychometric
instrument in the South African context (Heuchert et al., 2000; Meiring et al., 2005; Nel,
2008). Home language and language proficiency impact on responses to psychometric
instruments due to the eleven first languages spoken in South Africa amongst its many
cultures (Laher, 2010). Studies performed by Abrahams (1996) and Abrahams and Mauer
(1999) in South Africa, reveal that participants who took a test in English or Afrikaans,
neither of which was their first language, found difficulty in understanding common words
and the construction of sentences. Heaven and Pretorius (1998) investigated whether the
language descriptors of the FFM were adequate when used by a non-English-speaking group.
The instrument was translated into Afrikaans and Sotho; the participants’ native languages.
The traditional five factors were found for the Afrikaans-speaking group, but a different
pattern emerged for the Sotho-speaking group. The clearest factor to appear in the Sotho
sample was Emotional Stability.
1.10.3. Gender and Personality
Research into gender differences with regards to the personality structure of the FFM has
revealed that males are found to be higher on Assertiveness, while females score higher on
Anxiety and Tender-Mindedness (Croxford, 2009; Feingold, 1994; Laher, 2010). Costa,
Terracciano and McCrae (2001; Laher, 2010; Croxford, 2009) extended on Feingold’s (1994)
research by assessing all 30 traits across 26 cultures using the NEO-PI-R. The U.S and other
cultures revealed that females scored higher on Neuroticism, Agreeableness and
Extroversion. The U.S sample did not show significant differences in Conscientiousness and
Openness to Experience, while the other cultures found females to score higher than males on
these factors. This study was then repeated on 50 cultures using observer NEO-PI-R ratings
by McCrae et al.(2005) and the same results were found. However, Hyde (2005) argues that
these differences across gender are inflated and create misinformed stereotypes; thus
emphasising the importance of effect sizes.
27
Studies using the FFM have revealed a greater difference in gender in Western cultures than
African cultures, and thus may represent differences in personality traits between
individualistic and collectivistic cultures (Costa e al., 2001; McCrae et al., 2005; Laher, 2010;
Croxford, 2009). Costa et al. (2001) found that Zimbabweans and Black South Africans
reported little gender differences on the domains and facets of the FFM, while in Belgium
females were found to score higher on all domains. White female South Africans were also
found to be higher on all domains than males- differences were found to be weaker than for
the Belgian sample. McCrae et al. (2005), on the other hand, found no significant gender
differences in Nigeria, India, Botswana and Ethiopia; however in the United Kingdom
females were found to score higher than males on all domains.
Thus it is clear that South Africa is not a homogeneous group in terms of race, home
language and gender. Various differences are present across cultural groups and thus bias is
likely to occur in the utility of emic-etic instruments such as the SAPI and CPAI-2. This
study thus explored construct bias of the SAPI and CPAI-2 for race, home language and
gender. Construct bias was explored using ANOVAs at scale level.
1.11. Conclusion
Thus this research explored the collective dimension of personality using the South African
Personality Inventory (SAPI) and the Cross-cultural Personality Assessment Inventory-2
(CPAI-2) in a sample of South African first-year Psychology students from the University of
the Witwatersrand. This was done through an investigation into the internal consistency
reliability, construct validity and construct bias of the two instruments. Therefore this
research contributed to local and international research on personality theory and personality
assessment.
28
Chapter 2: Methods
2.1. Aims of the Study
This study aimed to explore the collective dimension of personality using the South African
Personality Inventory (SAPI) and the Cross-cultural Personality Assessment Inventory-2
(CPAI-2) in a sample of South African first-year Psychology students from the University of
the Witwatersrand. The internal consistency reliability, construct validity and construct bias
of the collective dimensions of the two instruments were investigated. Additionally thick
descriptions were obtained from two focus groups exploring issues around personality,the
collective dimension of personality, personality assessment, and the two instruments
themselves.
2.2. Rationale for this Study
This study contributed to both the local and international context in a number of waysincluding a contribution towards personality theory and assessment. In terms of personality
theories, the Five-Factor Model (FFM) is one of the most dominant models of personality and
is the most widely accepted as the standard framework of personality universally (McCrae &
Costa, 2008a). The NEO-PI-R is recommended as the measure of choice for operationalising
the FFM (Costa & McCrae, 1992; Church, 2000; McCrae & Costa, 2008a). However the
FFM has been criticised for its primarily Eurocentric individualistic focus on aspects of
personality as a result of its development in the English lexicon (Church, 2000; McCrae et al.,
2005). Personality assessment instruments, like the NEO-PI-R, have also been developed
primarily in relation to the FFM and therefore do not address the collectivist dimension of
personality. Such etic instruments are mainly in English, which is potentially problematic in
South Africa as this is not the first languagefor many South Africans (Wallis &Birt, 2003).
There have been strong cross-cultural claims pertaining to the universality of the FFM and
consequently the NEO-PI-R (McCrae & Costa, 1997; McCrae et al., 2005). However very
few studies have included African, Asian and Arabic cultures with regard to the NEO-PI-R
(Costa & McCrae, 1992) and so the FFM may be questionable in its applicability to these
29
cultures (Church, 2000; Cheung et al., 2003; McCrae et al., 2005; Paunonen& Ashton, 1998).
Hence there is a need for assessment instruments which do address this collectivist dimension
and consequently there is a need to validate and examine these instruments across cultures.
Theoretically this study also contributed to the etic-emic debate. Previous research has found
the use of imported-etic assessments in different cultures to be inadequate due to their
omission of indigenous culture-specific personality traits found in the emic approach (Laher,
2010; Yik& Bond, 1993). This inadequacy consequently results in validity issues with
regards to assessment across cultures and would prove to be most pertinent in a multi-racial
and multi-cultural country such as South Africa (Nel, 2008). It is believed that personality
structure may be universal, but that it is expressed in different ways in different cultures
(Meiring, 2006). Thus the use of the emic approach in South Africa would be suitable for its
rich cultural context when developing a personality assessment (Nel, 2008).
It has also been suggested that the West (Western Europe, North America, Australia) is
characterised by a focus on individualism; while the non-West (Asia, South America, Africa)
is characterised by collectivistic traits (Laher, 2010; Green et al., 2005). Etic measures, such
as the NEO-PI-R, are said to measure Western or individualistic dimensions of personality.
Thus the issue lies in that South African and Asian collectivist personality profiles would not
conform to Western individualist theories of personality. This is the premise on which the
SAPI and CPAI-2 were developed. Current research makes use of an etic-emic approach in
order to capture these indigenous cultural differences, while also allowing for comparisons
between these indigenous dimensions and other cultures (Saucier & Goldberg, 2001). The
two psychometric assessments were thus developed in order to measure the indigenous
dimensions of personality that were excluded from etic measures of personality, such as the
NEO-PI-R. These collective measures include Facilitating, Integrity, Relationship Harmony
and Soft-Heartedness (Nel, 2008), and Interpersonal Relatedness (Cheung, 2004)
respectively.
Linked to the etic-emic debate and of both theoretical and practical significance are the three
notions of culture, language and personality. These three notions create a powerful dynamic
30
that has a significant effect when assessing personality (Wallis &Birt, 2003). Language
comprehension and the system of meaning that is created, as well as the cultural nuances
found in Western instruments, may not be understood by all South Africans, thus cultural and
linguistic inappropriateness may be encountered. Studies reveal that etic measures of
personality used in a South African collectivist context exhibit bias in terms of cultural and
linguistic inappropriateness for this sample (Nel, 2008). Legislation has been developed in
order to regulate the use of fair and unbiased psychometric assessments. The Employment
Equity Act (Office of the President, 1998) regulates who will gain access to economic and
educational opportunities by stating that psychological testing is prohibited unless the
assessment being used is valid and reliable, applied fairly to all employees and is not biased
against any employee or group. Thus the merits of employing foreign or etic instruments
within South Africa have come into question. The exploration of issues such as validity,
reliability and bias in psychological instruments such as the SAPI and CPAI-2 will therefore
allow for the equal distribution of such opportunities in South Africa amongst its many
cultures and language groups (Sehlapelo& Terre Blanche, 1996; Laher, 2010).
Of particular local significance is the view of personality assessment as a means with which
to predict academic and job performance (Roodt&La Grange, 2001; Nel, 2008). This has
gained considerable ground over the last 10 years and is important as psychological
assessments including personality assessments, are pivotal in determining selection,
placement and management of employees (Sehlapelo& Terre Blanche, 1996). This
subsequently determines who will obtain access to economic and educational opportunities
(Owen &Taljaard, 1996 as cited in Van der Merwe, 2002; Nel, 2008; Laher, 2010).
This study thus aimed to assess reliability, validity and bias of the collective dimensions of
the SAPI and CPAI-2 in a South African context. This quantitative exploration was also
accompanied by a more qualitative analysis. It is believed that through this qualitative
exploration a more in depth analysis into the collective dimensions of the SAPI and CPAI-2
can take place. This will allow for progression towards more fair personality assessment in
South Africa, as outlined by the Employment Equity Act.
2.3. Research Questions
31
2.3.1. Quantitative Phase
Do the SAPI collectivist sub-clusters demonstrate adequate internal consistency reliability?
Do the CPAI-2 Interpersonal Relatedness scales demonstrate adequate internal consistency
reliability?
Do the SAPI collectivist sub-clusters have adequate construct validity?
Do the SAPI items demonstrate adequate construct validity?
Do the CPAI-2 Interpersonal Relatedness scales demonstrate adequate construct validity?
How do the SAPI collectivistic sub-clusters and the CPAI-2 Interpersonal Relatedness scales
load in a joint factor analysis and what are the implications of this for the collectivist
dimension of personality?
Does race influence responses on the SAPI?
Does race influence responses on the CPAI-2?
Does gender influence responses on the SAPI?
Does gender influence responses on the CPAI-2?
Does home language influence responses on the SAPI?
Does home language influence responses on the CPAI-2?
2.3.2. Hypotheses
32
There are no hypotheses for reliability and construct validity.
Race will influence responses on both the SAPI and CPAI-2, respectively.
Gender will influence responses on both the SAPI and CPAI-2, respectively.
Home language will influence responses on both the SAPI and CPAI-2, respectively.
2.3.3. Qualitative Phase
What are first-year psychology students’ perceptions of personality in South Africa?
What are first-year psychology students’ perceptions of a collectivist dimension to
personality in South Africa?
What are first-year psychology students’ perceptions of the instruments (SAPI and CPAI-2)
they completed?
2.4. Sample
2.4.1. Quantitative Phase
This study utilised data collected for a larger study that explored the collective dimension of
personality using the SAPI and CPAI-2 in South Africa. Non-probability convenience
sampling was used. The initial sample consisted of 516 students out of a possible 700 thus the
response rate was approximately 73.71%; however on examination 27 participants had to be
excluded from the sample as a result of largely incomplete data or too many missing items
across both the SAPI and CPAI-2 questionnaires. Participants with more than 20 missing
items were classified as invalid cases. Feedback from students during the collection phase
revealed that the questionnaire was felt to be too long, particularly for the SAPI, and that
items were repetitive which caused the students to get bored and lose interest. The final
sample thus consisted of 489first-year Psychology students at the University of the
Witwatersrand.
The Descriptive statistics presented in this section reveal that these students represent a
plethora of cultures, English proficiency, socioeconomic backgrounds, races and language
33
groups that represent South Africa. However, these students may not be representative of the
population in terms of education and age. It is acknowledged in literature that students are
generally not representative of all South African populations and thus the results may lack
generalisability (Rosenthal &Rosnow, 1991). Trends in psychometric research, however,
reveal that students are often used in exploratory studies and represent a relatively
homogenous sample, thus this sample is deemed to be appropriate. Student samples minimise
the effects of English proficiency and test wiseness (Abrahams &Mauer, 1999; Foxcroft,
1997), as both the SAPI and CPAI-2 were administered in English to students attending an
undergraduate degree at an English-medium university. Students are generally also familiar
with test-taking and their procedures.
It is evident from the Table 2.1 that 21.80% of the sample were male (n= 104) and 78.20%
were female (n= 373). Thus the sample in terms of gender revealed a female majority, with
females outnumbering males by nearly 3:1. The population groups represented in this sample
were White, Black, Indian and Coloured. 44.54% of the participants classified themselves as
White (n= 204), 36.24% as Black (n= 166), 13.10% as Indian (n= 60), and 6.11% as
Coloured (n= 28). First language groupings were distributed as follows: 3.63% of the sample
selected Afrikaans (n= 17) as their first language, 63.89% selected English (n= 299), 0.43%
selected IsiNdebele (n= 2), 3.42% indicated IsiXhosa (n= 16), 9.40% chose IsiZulu (n= 44),
4.91% selected Sepedi (n= 23), 2.78% selected Sesotho (n=13), 4.27% selected Setswana (n=
20), 1.92% indicated SiSwati (n= 9), 0.64% chose Tshivenda (n=3), 2.99% indicated
Xitsonga (n= 14), and 1.71% (n= 8) chose the Other option, specifying Chinese, amongst
others, as their first language. Therefore the majority of participants were first language
English speakers (n= 299), followed by IsiZulu (n= 44).
Table 2.1: Demographic statistics for the sample
Variable
Frequency
Percentage
Cumulative
34
Gender
Race
First/Home
Language
Education
Level
English
Reading Ability
Male
Female
Missing
White
Black
Indian
Coloured
Missing
Afrikaans
104
373
12
204
166
60
28
31
17
21.80
78.20
Percent
21.80
100.00
44.54
36.24
13.10
6.11
44.54
80.79
93.89
100.00
3.63
3.63
English
IsiNdebele
IsiXhosa
IsiZulu
Sepedi
Sesotho
Setswana
SiSwati
Tshivenda
Xitsonga
Other
Missing
Grade 9
299
2
16
44
23
13
20
9
3
14
8
21
0
63.89
0.43
3.42
9.40
4.91
2.78
4.27
1.92
0.64
2.99
1.71
67.52
67.95
71.37
80.77
85.68
88.46
92.74
94.66
95.30
98.29
100.00
0.00
0.00
Grade 12
Certificate
Diploma
Bachelors
Honours
Masters
Doctorate
Other
Missing
Very poor
317
4
7
115
1
0
0
4
41
1
70.76
0.89
1.56
25.67
0.22
0.00
0.00
0.89
70.76
71.65
73.21
98.88
99.11
99.11
99.11
100.00
0.21
0.21
Poor
Good
Very Good
Missing
4
179
282
23
0.86
38.41
60.52
1.07
39.48
100.00
Language proficiency is consistently identified as a factor affecting responses to
psychological tests, as presented in the literature. This is of particular relevance in South
Africa due to its multi-lingual and multi-cultural population (Nel, 2008). The University of
the Witwatersand is an English speaking university with a certain level of English proficiency
35
required by all students, thus such issues are deemed to be minimal. However to further
reduce this effect, students were asked to rate their English reading ability. The majority of
participants indicated a ‘Good’ to ‘Very Good’ English reading ability (see Table 2.1); with
38.41% of the sample selecting ‘Good’ (n= 179) and 60.52% choosing ‘Very good’ (n=282).
1.07% of participants rated their English reading ability to be ‘Very poor’ to ‘Poor’; with 1
participant (0.21%) selecting the ‘Very poor’ rating.
Previous research has indicated that participants’ education level is believed to have a
possible effect on the ability of participants to understand the sometimes sophisticated
English utilised in some psychometric tests. Education is believed to allow participants a
greater opportunity to understand what the items require of them (Branco e Silva &Laher,
2010). This study consists of first-year students, thus reducing these potential issues.
However, to further minimise the possible effects of education, participants were asked to
indicate their highest level of education, these results, as per the Table 2.1, included: 70.76%
of the sample indicated Grade 12 (n= 317) as their highest education level, 0.89% selected a
Certificate (n= 4), 1.56% as a Diploma (n= 7), 25.67% selected a Bachelors (n= 115), 0.22%
with Honours (n= 1), and 0.89% (n= 4) chose the Other option, but did not specify their
grouping.
Table 2.2: Demographic statistics for age and English reading ability
Variable
Mean
Min
Max
Range
N
19.0625000
Standard
Deviation
2.0759015
Age
Reading
Ability
17
44
27
400
3.5922747
0.5216438
1
4
3
466
The students in the sample ranged from 17 to 44 years of age, with a mean of 19.1 (SD=
2.08), as displayed in Table 2.2. The mean English reading ability of participants in this study
was 3.6 (SD=0.52), indicating that most participants revealed a ‘Good’ or ‘Very good’
English reading ability as presented in Table 2.2. and in the demographic information Table
2.1.
2.4.2. Qualitative Phase
36
While completing the questionnaire, students were asked to volunteer to participate in focus
groups discussing personality and personality assessment in the South African context, in
order to further explore the two instruments completed- the SAPI and CPAI-2. Participants
were invited to one of two focus group sessions which were held in the second quarter of the
university year. A non-probability, conveniencesample of 9 individuals for focus group 1,
and 10 individuals for focus group 2, were obtained from the original sample who
volunteered to complete the questionnaire. Thus the final sample consisted of 19 first-year
Psychology students at the University of the Witwatersrand for both focus groups. Students
self-selected into the two focus groups based on what was suitable for them.
Table 2.3: Descriptive statistics for gender and race
Variable
Gender
Race
Male
Female
Black
White
Indian
Frequency
Percentage
4
15
7
11
1
21.05
78.95
36.84
57.90
5.26
Cumulative
Percent
21.05
100.00
36.84
94.74
100.00
The 19 participants consisted of 15 females (78.95%) and 4 males (21.05%); thus revealing a
female majority of nearly 4:1 (see Table 2.3). The 2 focus groups consisted of a White
majority of 57.90% (n= 11), followed by 36.84% Blacks (n= 7), and 5.26% Indians (n= 1).
2.5. Instruments
2.5.1. Quantitative Phase
A questionnaire 1 which incorporated a demographics section, the social-relational items of
the SAPI and the interpersonal relatedness items of the CPAI-2 was used in this study. Two
versions of the questionnaire were utilised; one with the SAPI first followed by the CPAI-2,
and the other with the CPAI-2 first followed by the SAPI. This counterbalancing technique
was used in order to ensure that adequate and relatively equal numbers of participants
completed the SAPI and CPAI-2.
1
See Appendix A for sample questionnaire.
37
2.5.1.1. Demographic section
Students were asked to provide their demographic information; namely age, gender, highest
level of education, self-evaluated English reading ability, race and home language. The
demographic information related to race, home language and gender was used for analysis;
while the remaining information obtained in the questionnaire was used for descriptive
purposes only and allowed for a better understanding of the people that responded to the
questionnaire.
2.5.1.2. The SAPI
The social-relational cluster of the SAPI constitutes the clusters of Facilitating, Integrity,
Relationship Harmony and Soft-Heartedness (Nel, 2008). The cluster Soft-Heartedness is
comprised of the sub-clusters Amiability, Egoism, Gratefulness, Hostility, Empathy and
Active Support (Nel, 2008). Amiability is the quality of being kind trusting and pleasant.
Egoism focuses on one’s selfish needs and desires, and describes the quality of being greedy
and jealous. Gratefulness defines one’s tendency to be appreciative. The sub-cluster hostility
encompasses one’s inclination to be aggressive and to enjoy the physical and emotional abuse
of others. Empathy is the quality of being compassionate and caring for others needs and
feelings; while Active Support is the quality of being generous, and being actively involved
in the well-being of others (Nel, 2008).
The cluster Integrity is compiled from the sub-clusters Integrity and Fairness. Integrity is
one’s inclination to be honest, pure, trustworthy and responsible; whereas Fairness is one’s
inclination to accept and treat people equally (Nel, 2008). The cluster Relationship Harmony
is compiled of Approachability, Conflict-seeking, Interpersonal Relatedness and
Meddlesome. Approachability is the ability to be approachable and accessible for
interpersonal interactions with other; whereas Conflict-Seeking is one’s tendency to provoke
and disrupt others and cause conflicts. Interpersonal Relatedness is the ability to be peaceful,
forgiving, constructive and cooperative in one’s relationships in order to maintain them.
Meddlesome is the inclination to interfere in the lives of others (Nel, 2008).
38
The final cluster Facilitating is made up of the sub-clusters Guidance and Encouraging
Others. Guidance is one’s ability to be a role model by guiding and teaching others about
right and wrong through personal example. Encouraging Others is the quality one possesses
that allows them to encourage others to dream and reach their potential (Nel, 2008).
The original social-relational scales incorporating these clusters and sub-clusters contained a
total of 265 items. However, data from the pilot studies indicated that not all the sub-clusters
or facets could be replicated; while some clusters were found to be completely new- such as
Harmony Maintenance and Harmony Breach of the Relationship Harmony cluster(V.H.
Valchev, personal communication, September 12, 2011).Specifically, the Integrity subclusters were not replicated in the pilot studies; therefore this cluster does not include subclusters, revealing that it may not be a valid factor on its own. Therefore, because this
instrument is still under construction, items were chosen for this questionnaire that came as
close as possible to representing the sub-clusters and the new emergent sub-clusters were
added to the questionnaire- this included the addition of a Social Desirability cluster (V.H.
Valchev, personal communication, September 12, 2011).This cluster incorporates items from
a number of existing social desirability scales; however mostly from the Marlowe-Crowne
Social Desirability Scale (MCSDS) and Paulhus’s Balanced Inventory of Desirable
Responding (BIDR) as these are considered the most prominent. Consequently, the final subclusters of this questionnaire for the five clusters included- Egoism, Hostility, Empathy, I
help others to succeed, Active Support, Guidance, Encouraging Others, Integrity,
Approachability, Harmony Maintenance, Harmony Breach, Social Desirability, Impression
Management and Self-deception (V.H. Valchev, personal communication, September 12,
2011).
The pilot data reported the following Cronbach’s alpha coefficients: Facilitating- .97 (55
items), Integrity- .61 (34 items), Relationship Harmony- .80 (51 items) and .81 (48 items),
Soft-Heartedness- .86 (59 items) and .89 (78 items). Relationship Harmony and SoftHeartedness contained two data sets with subsets of the scales. The pilots included larger
scales than the ones used in this study. The Cronbach’s alpha coefficients presented above
39
refer to the restricted sets of items used in the present study (V.H. Valchev, personal
communication, April 6, 2011).
With regards to the factorial structure, there is evidence for a relatively unifactorial structure
for Facilitating and Integrity- although there is a possibility for weaker second factors. The
factors of Harmony Breach, Harmony Maintenance and Arrogance of Relationship Harmony;
as well as Respect, Hostility, Egoism, Active Support, Greed, Empathy and Distrust for SoftHeartedness are largely supportive of the expected structures. It is important to note,
however, that these validity measures were obtained when the scales were administered
separately in the pilot study and would thus be expected to change in this study. (V.H.
Valchev, personal communication, April 6, 2011).
2.5.1.3. The CPAI-2
The CPAI-2 Form B is a self-report instrument consisting of 341 items. The CPAI-2 is a
suitable questionnaire for individuals in the age range of 18 – 70 with a Grade six as a
minimum level of education. It consists of four factors and 28 personality scales; namely
Social potency (8 scales), Dependability (9 scales), Accommodation (5 scales) and
Interpersonal Relatedness (6 scales). The CPAI-2 is available in 4 languages; including
Chinese, Korean, Japanese and English (Cheung, 2004). The English version was utilised in
this study. A 5-point rating scale was used for this instrument instead of the original true-false
format in order to achieve uniformity in rating responses between the SAPI and CPAI-2 for
this study (V.H. Valchev, personal communication, December 14, 2011).
This study aimed to explore the collective dimensions of the CPAI-2; thus only the 74
Interpersonal Relatedness items were used in this study. The Interpersonal Relatedness factor
incorporates 6 scales namely; Traditionalismvs Modernity, Ren Qing (Relationship
Orientation), Social Sensitivity, Discipline, Harmony and Thrift vs Extravagance.
Traditionalism vs Modernity measures attitudes toward Chinese beliefs and values in areas
such as family relationships, filial piety, rituals and chastity. This scale measures the degree
to which individuals are obedient, conservative, protect old customs and oppose new ideas; as
40
opposed to individuals that challenge traditional concepts and superstition, and advocate
democracy and personal freedom (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001).
Ren Qing or Relationship Orientation measures an individual’s adherence to cultural norms.
This measures the degree to which individuals strengthen their relationships, change their
principles for the needs of others and help others strengthen their social ties; or the degree to
which they are disinterested in social customs, and avoid interactions and the exchange of
favours with others (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001).
Social Sensitivity measures the extent to which individuals are sensitive to others;
empathetic, understanding, willing to listen and communicate, and are easily approachable
(Cheung, 2006; Cheung et al., 2001; Cheung et al., 2008). Discipline refers to the extent to
which individuals are rigid and disciplined, rather than flexible and adaptable. High scores
indicate individuals who are stubborn and organised, who dislike unpredictable events and
like convention. Low scores indicate individuals who adapt to changes in the environment
and the situation and discard formality (Cheung, 2006; Cheung et al., 2001).
Harmony is the degree to which an individual is content and possesses inner harmony and
peace of mind. High scores describe individuals who are even-tempered, peaceful and try not
to offend others. These are considered virtues, as opposed to low scores that describe
individuals who are confrontational, insensitive and welcome conflict (Cheung, 2006;
Cheung et al., 1996; Cheung et al., 2001). The Thrift vs Extravagance scale describes one’s
carefulness in spending. High scores describe thrifty individuals who maintain a simple life,
and are careful in their expenditure of money and resources. Low scores describe extravagant
individuals who spend a lot of money and flaunt their wealth (Cheung, 2006; Cheung et al.,
1996; Cheung et al., 2001).
This version of the CPAI was validated in a sample of 675 Singaporean Chinese and 144
Caucasian American students at a midwestern university, where similar factor structures were
obtained in both samples. Cheung (2004) reports a median Cronbach Alpha coefficient of .63
for the personality scales of the CPAI. Cheung et al. (2008) reports a Cronbach’s alpha of .70
for the Interpersonal Relatedness factor. The Chinese version of the CPAI-2 was also
41
validated in a Chinese sample and a similar result was obtained (Cheung, et al., 2003). The
Chinese version was standardised in a sample of 1911 adults from Mainland China and Hong
Kong (Cheung, Cheung, Howard & Lim, 2006; Cheung et al., 2008). Alpha coefficients for
this sample, on the 28 personality scales, range from .80 to .51 (Cheung et al., 2008). No
reliability coefficients were available for the factors. Evidence for the construct validity of the
CPAI-2 has been cited by Cheung et al. (2006); however the Openness and Interpersonal
Relatedness dimensions are still being debated (see Cheung et al., 2008). The CPAI-2 is a
fairly new instrument, thus information on its psychometric properties is still being compiled
(Laher, 2010).
2.5.2. Qualitative Phase
The focus group questions 2 consisted of 7questions probing the participants’ perceptions of
personality in South Africa, as well as their perceptions of the SAPI and CPAI-2. The focus
group questions were developed after an extensive analysis of the literature review and in
conjunction with my supervisor. The first thing to do after the development of focus group
questions is to pilot it (Creswell, 2009). This was done by sending the focus group questions
to three of the first-year Psychology students. Their feedback was taken into account and the
focus group questions changed accordingly.
2.6. Procedure
2.6.1. Quantitative Phase
Students were approached during lecture times to complete the questionnaire. This took place
after permission was granted from the course coordinator and head of Psychology for the
first-year Psychology students. A sealed box was placed at the front of the lecture room and
students were asked to place the completed questionnaires and answer sheets 3 into the box.
Once questionnaires were received and the data captured the relevant statistical analyses were
conducted.
2
See Appendix C for focus group questions.
See Appendix A for sample SAPI answer sheet.
3
42
The questionnaires were printed and distributed as part of a larger study exploring the
collective dimension of the SAPI and the CPAI-2 in South Africa. I participated in this
collection and distribution in order to acquire the use of the data for my Masters dissertation.
My supervisor and I took on the responsibility of ensuring that a questionnaire was given to
each of the possible participants from the three Psychology first-year classes. Participants
were given approximately an hour to an hour and a half to complete and return the
questionnaire. My supervisor and I provided the participants with a sealed box. The box was
placed in the front of the lecture theatre for all participants to see. Once the questionnaires
were completed the participants placed them back into the box so as to secure the anonymity
of the instrument.
2.6.2. Qualitative Phase
After the completion of the questionnaires, participants were informed that a focus group
would be taking place at a later stage in order to obtain more information about the
participants’ perceptions of the questionnaire. Participants were notified that if they wished to
take place in a focus group they had to provide their name, cell phone number and email
address on a piece of paper that was circulated around the classroom by my supervisor and
myself. These details were then used in order for me to contact the participant to set a time
for the focus groups to take place. This was done in all three classes so as to conduct two
focus groups to discuss participants’ perceptions of personality in South Africa and their
perceptions of the SAPI and CPAI-2.
There were two separate time slots to choose from and each student attended one focus group
at their convenience. I provided the volunteers with details regarding the on-campus venue
for each focus group via email and SMS. On the day of each focus group I provided the
participants with a Participant Information Sheet and Informed Consent Forms 4 for
participation and for video-recording. Participants were then asked to read through these
sheets and sign were necessary. Signed and returned forms were regarded as permission and
4
See Appendix B for Participant Information Sheet and Informed Consent Forms.
43
consent from the volunteer to advance with the focus group procedures. Each focus group
participant was asked to detach and keep the Participant Information Sheet for their
information and interest. The researcher then asked the focus group questions while the
participants were recorded by a video tape recorder. The participants’ responses were
transcribed and thereafter analysed.
2.7. Research Design
The strategy used in this study was both quantitative and qualitative in nature. The research
design thus produced both statistical evidence and rich, descriptive accounts regarding a
multicultural South African perspective on personality assessment using the SAPI and CPAI2.
2.7.1. Quantitative Design
The quantitative section of the study used a non-experimental cross-sectional design. It is
cross-sectional as students completed the questionnaire at one point in time in a natural
setting. There was no manipulation of the independent variable, there were no control
measures and the volunteers were not randomised. Therefore it does not meet the
requirements for an experimental design and is thus non-experimental (Larson & Buss, 2005;
Rosenthal &Rosnow, 1991).
2.7.2. Qualitative Design
The qualitative section of this study utilised focus groups. Qualitative research designs are
believed to describe an individual’s lived human experience (Babbie& Mouton,
2004;Creswell, 2008). The purpose of such a design is to obtain an in-depth and
comprehensive understanding of the research topic through resource intensive methods.
Furthermore qualitative research aims to report participants’ views and individual
experiences in a particular context (Willig & Rogers, 2008). Qualitative research thus strives
to explore and understand the different viewpoints and experiences of the participants and is
thus better able to capture the diverse and complex meanings of participants’ subjective
44
experiences in their socio-cultural context (Willig & Rogers, 2008). Therefore data is not
presented in a reductionistic, static or decontextualised manner. Participants’ thoughts,
feelings and actions are acknowledged in an attempt to make sense of them (Willig & Rogers,
2008).Thus the utilisation of focus groups allowed me to obtain a better understanding of the
volunteers’ perceptions of personality in South Africa, as well as their perceptions of the
collective personality dimensions of the SAPI and CPAI-2. I facilitated the conversation to
relevant topics using appropriate questions which allowed me to participate in a descriptive
open informal discussion with volunteers of the focus group.
This study thus aimed to explore the collective dimension of personality of the SAPI and
CPAI-2 using both quantitative and qualitative methods. The advantage of this dual focus is
that it fulfils both the quantitative and the qualitative exploratory demands of evaluating the
collective dimension of personality. This thus allowed me to obtain an approximate
numerical index of subjective experience (Willig& Rogers, 2008). Furthermore, allowing for
a more comprehensive and in depth exploration, and consequently may result in more
conclusive evidence, as well as a more complete description of the collective dimension of
personality (Willig& Rogers, 2008). Reichardt and Rallis (1994; as cited in Willig & Rogers,
2008) reveal that the greatest advantage of a mixed method study is that quantitative and
qualitative approaches are expert at critiquing the methods and conclusions of the other
approach.Qualitative research therefore aims to question concepts and examine exceptions
and suppressed meanings, and as a result quantitative findings may need to be qualified
(Barbour, 2000; Willig & Rogers, 2008).
2.8. Ethical Considerations
2.8.1. Quantitative Phase
45
Participation in the research study was voluntary and informed consent was obtained.
Confidentiality and anonymity was maintained for the questionnaires in that no identifying
data was requested from participants. All volunteers were fully briefed about the research in a
cover letter attached to the questionnaire. The cover letter briefly outlined the purpose of the
study and the time that was required to complete the questionnaire. It provided a statement of
anonymity and confidentiality. If the volunteer wished to withdraw at any stage during the
introduction to the questionnaire, the volunteer had the right to do so and would not be
discriminated against in any way. A completed and returned questionnaire was regarded as
permission and consent from the volunteer for me to utilise the data. General feedback will be
available after completion of the questionnaire; however I will not be able to give feedback to
individual volunteers about their responses on the SAPI and CPAI-2 since the responses are
anonymous. Feedback on general results will be available 6-9 months after completion of the
questionnaire in a one to two page summary sheet which will be available on request from me
via email and will be placed on the first-year psychology notice board outside the psychology
office. Participants were informed that the study may be written up for publication in a
journal or presented at a conference. For this reason raw data will be stored in a locked
cupboard on campus for a period of up to three years and then destroyed. Participants were
asked to detach and keep the cover letter.
2.8.2. Qualitative Phase
All volunteers were fully briefed about the focus group in a cover letter before providing their
informed consent to participate. This included the nature of the study and the time
commitment for the focus group. Volunteers were informed that at any stage during the focus
group, the volunteer had the right to withdraw and would not be discriminated against in any
way. I obtained written informed consent from the volunteers about the focus group by way
of two consent forms which were given to participants on the day of the focus group. One
form obtained consent for the focus group to take place, while the other form obtained
consent for the video recording. Any responses that are discussed or quoted in the final
research dissertation were coded using pseudonyms e.g.F1, J3 to ensure participant
confidentiality. All participants were informed that my supervisor and I will be the only
people to have access to their responses, therefore confidentiality is ensured. Feedback on
general results will be available 6-9 months after completion of the questionnaire in a one to
46
two page summary sheet which will be available on request from me via email and will be
placed on the first-year psychology notice board outside the psychology office. Participants
were informed that the study may be written up for publication in a journal or presented at a
conference. For this reason raw data will be stored in a locked cupboard on campus for a
period of up to three years and then destroyed. Participants were asked to detach and keep the
cover letter.
2.9. Data Analysis
2.9.1. Quantitative Phase
Descriptive statistics were used for analysis in this study. This involved computing the data to
provide frequencies for the nominal variables under study which included; gender, race
education level and home language. Means, standard deviations, and minimum and maximum
scores were established for the variables like age and self-evaluated English reading ability
that are interval. Skewness coefficients were established for SAPI and CPAI-2 scales in order
to determine if they are normally distributed. This then allowed for the use of parametric
tests; provided that the assumptions of random independent sampling, interval dependent
variable and homogeneity of variance are met (Huck, 2009).
2.9.1.1. Internal Consistency Reliability
The Cronbach Alpha reliability coefficients were used to determine the internal consistency
reliability of the SAPI and CPAI-2. The purpose of this measure is to calculate inter-item
correlations in order to assess whether different parts of the instrument measure the same
construct (Murphy &Davidshofer, 2005).
2.9.1.2. Construct Validity
Exploratory factor analysis with Varimax rotation was used on the responses to the SAPI and
CPAI-2. The aim of exploratory factor analysis is to explore the field and determine the main
47
constructs and dimensions within this field. This method is a powerful tool in clarifying
determiners and associated variables of the personality factors under investigation. In this
way many variables undergo factor analysis in order to determine what variables load on the
relevant factors (Kline, 1994). “The goal of Varimax rotation is to simplify the columns of
the unrotated factor-loading matrix” (Pett, Lackey & Sullivan, 2003, p. 141). In order to
obtain this goal, Varimax maximises the variances of the loadings within the factors, as well
as between high and low loadings on a particular factor, thereby facilitating interpretation
(Pett et al., 2003). Exploratory factor analysis with Varimax rotation was also used in this
study as it was the method used by Cheung and colleagues (2003) and currently adopted and
accepted in studies of the FFM (McCrae et al., 2005).
2.9.1.3. Construct Bias
Construct bias was assessed through the use of ANOVAs in this study. ANOVAs were used
at scale level to assess construct bias. This statistical technique was used in analysing the
SAPI and CPAI – 2 collective dimension scale scores, respectively in relation to the variables
of race, gender and home language. This test was used in order to determine whether a
difference exists between this demographic information on the collective dimension scales of
each instrument. Parametric one-way ANOVAs were used provided that the assumptions for
a parametric test are met- dependent variables are normally distributed and are at least
interval, and the condition of homogeneity of variance is fulfilled. If not, Kruskal-Wallis oneway ANOVAs were used. The scores obtained for the SAPI and CPAI-2 are classified as
interval data. Normality was examined using the skewness coefficient, as discussed above.
Levene’s test for homogeneity of variance was used in order to assess whether the variances
were equal (Huck, 2009).
Furthermore, post hoc testing was not necessary in this study due to the comparison of only
two groups; however a comparison of means for parametric significant results was
conducted, indicating the direction of these results. Cohen’s d was utilised to calculate effect
sizes for the significant results of the parametric ANOVAs (Huck, 2009). Cut-offs suggested
for these effect sizes include; .20 for a small effect size, .50 for a moderate effect size and .80
for a large effect size (Huck, 2009). Cliff’s d could not be calculated for the non-parametric
48
Kruskal-Wallis one-way ANOVA significant results in this study, due to problems
encountered with the Cliff’s d formula.
2.9.2. Qualitative Phase
Thematic content analysis was used to analyse the transcriptions of the focus group answers.
This method aims to identify, analyse and report themes within data by organising and
describing the data set in detail (Boyatzis, 1998; Braun & Clarke, 2006). This process allows
for the emergence of themes, as well as the identification, selection and report of themes by
the researcher (Braun & Clarke, 2006;Taylor & Ussher, 2001). Thematic content analysis
takes place in a sequence of steps. The first step is for the researcher to familiarise
himself/herself with the data by reading and re-reading the data while noting down some
initial ideas. The second step involves the coding of interesting features of the data and the
collection of all the relevant data for each code. Next, codes must be transformed into
potential themes and all relevant data for that theme must be gathered. A thematic map on the
analysis must be developed to ensure that themes work in relation to coded extracts and the
entire data set. Then themes need to be named and clearly defined and finally an analysis of
the themes needs to be given (Braun & Clarke, 2006). I followed the steps outlined above by
Braun and Clarke in my qualitative analysis (2006).
2.10. Self-Reflexivity
Qualitative analyses, such as thematic content analysis, allow the researcher a voice in which
information that supports the researcher’s argument can be selected and edited (Fine, 2002).
Reflexivity is a methodological tool that better represents, legitimises, or calls into question a
researcher’s data (Pillow, 2003). It is important for the researcher to attempt to be selfreflexive in order to be aware of preconceptions and expectations in terms of their research.
In this way subjectivity may be minimised as an issue. Thus, in this study I continuously
partook in self-analysis and political awareness, so as to be conscious of how my selflocation (across for example, gender, race, class, sexuality, ethnicity, nationality), position,
and interests influenced the research process (Pillow, 2003); as well as the ways in which
they may have been perceived by others (Callaway, 1992, as cited in Pillow, 2003).
49
In the interest of self-reflexivity, I acknowledge that this research is a study exploring the
collective dimensions of the SAPI and CPAI-2 in South Africa done by a South African
Industrial Psychology student. Thus I must be made aware of my possible preconceptions and
expectations about both the participants, and the nature of the data to be obtained. These
preconceptions and expectations stem from the results of previous research I have read, as
well as from being exposed to psychological tests as a result of being a psychology student. I
can also be seen as having a vested interest in the chosen topic, as I am expecting a certain
result and this can affect how I phrase my focus group questions and how I analyse the data
obtained. I also want to obtain a good mark for my Masters project and thus I am hoping to
obtain rich results. Although these facts cannot be ignored, I continuously self-reflected on
my role as researcher and continuously attempted to obtain as neutral a stance as possible
throughout the study. It should also be considered that as a researcher, the need to search for
information and increase understanding plays an equally significant role. Hence we juxtapose
these roles and by virtue of acknowledging our own biases and limitations hope to present as
true a reflection of participants’ responses as possible.
2.11. Conclusion
This study therefore sought to investigate the collective dimensions of the SAPI and the
CPAI-2 in the South African context, by exploring issues of construct validity, internal
consistency reliability and construct bias. It is hoped that through this research the SAPI
moves closer to being considered a valid and reliable personality structure and personality
assessment in South Africa.
50
Chapter 3: Results
3.1. Introduction
This chapter presents all the results obtained in this study. This study focused on exploring
the collective dimension of personality, using the SAPI and CPAI-2 in a South African
sample,through a quantitative and qualitative analysisthat are presented below.
Descriptive statistics for the sample in terms of gender, race and home language are initially
provided; followed by descriptive statistics for the SAPI and CPAI-2. Following from this,
results relating to reliability in terms of the internal consistency of the SAPI and CPAI-2 will
be presented. This chapter will then address construct validity by presenting the results of the
exploratory factor analyses conducted in terms of the SAPI and CPAI-2. Finally results
pertaining to construct bias will be presented for the two instruments, using ANOVAs at
scale level in relation to gender, race and home language.
The qualitative thematic content analysis explored the collective dimension of personality in
the South African context,by exploring participants’ perceptions of personality in South
Africa, as well as their perceptions of the SAPI and CPAI-2, by virtue of seven themes. The
results for this analysis are also presented below.
3.2.Descriptive Statistics: Demographic Information
Three demographic variables were of particular relevance in this study; namely gender, race
and homelanguage. The subgroups of race and home language have been collapsed due to
issues of sample size relating to groups being significantly unequal. It is believed that such
issues would hamper further statistical analyses and thus were collapsed in an attempt to
overcome the potential problems. Race was originally assessed in terms of Blacks, Whites,
Coloureds and Indians- these groups have henceforth been collapsed into the groupings
White and non-White. Home language was originally assessed looking at Afrikaans, English,
IsiNdebele, IsiXhosa, Sepedi, Sesotho, Setswana, SiSwati, Tshivenda, Xitsonga and Other.
Previous research in the South African context has suggested the effect of one’s home
51
language on test performance (see Foxcroft, 2004; Franklin, 2009).Thus, due to the utility of
the SAPI and CPAI-2 in this assessment in English- home language will be collapsed into the
groupings English first language speakers and English second language speakers. These
collapsed groupings, although not homogenous, will therefore allow for an exploration of
bias in the South African context.
Table 3.1: Frequencies for gender, race and home language
Variable
Gender
Race
Home Language
Frequency
Percentage
Male
104
21.80
Female
373
78.20
White
204
44.54
Non-White
254
55.46
English first language
299
63.89
English second language
169
36.11
From the Table 3.1 it is evident that a large majority of the sample were female (78.20%).
The sample also consisted of 44.54% Whites and 55.46% non-whites; while the majority of
participants (63.89%) were English first language speakers.
3.3.Descriptive Statistics: SAPI
Table 3.2 presents the means, standard deviations, minimum and maximum values, and the
skewness coefficients for the clusters and sub-clusters of the SAPI. From this table it is clear
that all the variables are normally distributed as the skewness coefficients do not exceed +1
or -1 (Huck, 2009). Soft-Heartedness has a mean of 325.23 (SD = 22.33) and scores ranging
between 244 and 408. Facilitating has a mean of 200.89, with a standard deviation of 25.71,
and scores ranging between 102 and 271. Integrity has a mean of 115.1 (SD = 7.97) and
scores ranging between 92 and 143. Relationship Harmony has a mean of 248.58 with a
52
standard deviation of 15.78 and scores ranging between 201 and 294. Lastly Social
Desirability has a mean of 84.38 (SD = 7.33) and scores ranging from 59 to 107.
Table 3.2: Descriptive Statistics for the SAPI
Cluster/
Mean
SD
Minimum
Maximum
Skewness
325.23
22.33
244
408
0.105
26.02
86.69
127.12
81.73
3.69
5.06
17.14
13.32
9.15
0.83
14
41
72
36
1
42
150
158
105
5
0.307
0.309
-0.209
-0.187
-0.621
200.89
161.44
39.45
25.71
21.62
5.22
102
83
19
271
222
50
-0.027
0.028
-0.374
115.1
248.58
7.97
15.78
92
201
143
294
0.298
0.245
37.20
171.42
6.52
17.39
21
113
64
218
0.640
0.126
39.96
8.48
21
67
0.289
84.38
7.33
59
107
0.078
39.45
4.12
25
53
0.216
21.23
2.82
9
29
-0.387
23.70
3.14
14
35
0.134
Sub-cluster
Soft
Heartedness
Egoism
Hostility
Empathy
Active Support
I help others to
succeed
Facilitating
Guidance
Encouraging
Others
Integrity
Relationship
Harmony
Approachability
Harmony
Maintenance
Harmony
Breach
Social
Desirability
Social
Desirability
Impression
Management
Self-deception
Table 3.2 indicates that all the Soft-Heartedness sub-clusters, except for Egoism (M = 26.02)
and ‘I help others to succeed’ (M = 3.69), have means of between 81 and 128and standard
deviations between 9.15 and 17.14. Minimum and maxim scoresprovide a wide range from 1
53
to 158; with minimum scores ranging from between 36 to 72 and maximum scores ranging
from 105 to 158 for Hostility, Empathy and Active Support. The range for ‘I help others to
succeed’ is between 1 and 5, while the range for Egoism is between 14 and 42.
The Facilitating sub-clusters provided in the table suggest means of 39.45 and 161.44 and
standard deviations of 5.22 and 21.62, for Encouraging Others and Guidance respectively.
The minimum and maximum range for Encouraging Others is between 19 and 50, while the
range for Guidance is that of 83 to 222.
The Relationship Harmony sub-clusters, as in Table 3.2, provide means of between 37 and 40
with standard deviations of between 6.52 and 8.48, with the exception of Harmony
Maintenance; this sub-cluster has a mean of 171.42 with a standard deviation of 17.39. The
sub-clusters Approachability and Harmony Breach have minimum and maximum scores
ranging from 21 to 64 and 21 to 67 respectively. Harmony Maintenance, on the other hand,
has minimum and maximum scores ranging from 113 to 218.
The Social Desirability sub-clusters: Impression Management and Self-deception have means
of between 21 and 24, with standard deviations of between 2.82 and 3.14. The sub-cluster
Social Desirabilityhas a mean of 39.45 and a standard deviation of 4.12. Minimum and
maximum scores for the sub-clusters of Social Desirability provide quite a variety with the
minimum score for Social Desirability being 25, Impression Management 9 and Selfdeception 14. Maximum scores also indicate a variation with the maximum score for Social
Desirability being that of 53, Impression Management 29 and Self-deception 35.
3.4.Descriptive Statistics: CPAI-2
Table 3.3 presents the means, standard deviations, minimum and maximum values, and the
skewness coefficients for the factor and scales of the CPAI-2.This table indicates that all the
variables are normally distributed, as the skewness coefficients do not exceed +1 or -1 (Huck,
54
2009). Interpersonal Relatedness has a mean of 241.75 (SD = 18.68) and scores ranging from
184 to 301.
Table 3.3: Descriptive Statistics for the CPAI-2
Scale
Mean
SD
Minimum
Maximum
Skewness
Interpersonal
Relatedness
Traditionalismvs
Modernity
Ren Qing
Social
Sensitivity
Discipline
Harmony
Thriftvs
Extravagance
241.75
18.68
184
301
0.050
39.61
6.84
21
57
-0.002
42.03
42.81
4.76
4.7
26
21
58
56
0.034
-0.228
35
50.66
31.64
5.05
5.93
4.3
14
34
18
48
67
45
-0.061
0.04
0.110
The scales of the Interpersonal Relatedness factor, as indicated in Table 3.3, provide a
variation in means with Traditionalismvs Modernity being that of 39.61, Discipline of 35,
Harmony of 50.66 and Thrift vs Extravagance of 31.64. Ren Qing and Social Sensitivity,
however, have similar means with that of 42.03 and 42.81 respectively. Standard deviations
for these scales range from between 4.3 to 6.84; with Ren Qing and Social Sensitivity (4.76
and 4.7), as well as Discipline and Harmony (5.05 and 5.93)indicating similar standard
deviations. A variation in minimum and maximum scores also exists for the Interpersonal
Relatedness scales; Traditionalism vs Modernity having a minimum score of 21, Ren Qing a
score of 26, Social Sensitivity of 21, Discipline of 14, Harmony of 34 and Thrift vs
Extravagance of 18. A variation similarly exists for maximum scores of this factor;
Traditionalismvs Modernity having a maximum score of 57, Ren Qing a score of 58, Social
Sensitivity of 56, Discipline of 48, Harmony of 67 and Thrift vs Extravagance of 45.
55
3.5. Reliability: SAPI and CPAI-2
Internal consistency reliabilities were examined for both instruments- the SAPI and the
CPAI-2. The results for this reliability analyses are presented below.
3.5.1. Internal Consistency Reliability: SAPI and CPAI-2
Cronbach Alpha reliability coefficients for each cluster and sub-cluster of the SAPI and the
Interpersonal Relatedness factor and scales of the CPAI-2 are provided in Table 3.4. This
table presents reliability coefficients of .86, .96, .63 and .81 for the clusters of the SAPI: SoftHeartedness, Facilitating, Integrity and Relationship Harmony, respectively; as well as .79 for
the Interpersonal Relatedness factor of the CPAI-2.
Table 3.4: Internal consistency reliability coefficients for the SAPI and CPAI-2
SAPI
Soft Heartedness
α
0.86
Egoism
0.62
Hostility
Empathy
Active Support
I help others to
succeed
Facilitating
0.90
0.92
0.90
Guidance
Encouraging Others
Integrity
Relationship
Harmony
Approachability
Harmony
Maintenance
Harmony Breach
Social Desirability
Social Desirability
Impression
Management
Self-deception
0.96
0.86
0.63
0.81
0.96
0.71
0.93
0.82
0.53
0.33
0.17
0.20
CPAI-2
Interpersonal
Relatedness
Traditionalismvs
Modernity
Ren Qing
Social Sensitivity
Discipline
Harmony
α
0.79
Thriftvs
Extravagance
0.36
0.66
0.44
0.60
0.57
0.68
56
Reliability coefficients for the Soft-Heartedness sub-clusters range from .62 to .92; with an
average Cronbach alpha of .84. The median reliability coefficient for the Soft-Heartedness
sub-clusters is .90; internal consistency reliability coefficients for these sub-clusters are all
.90 and above with the exception of Egoism (α= .62). No reliability was found for the ‘I help
others to succeed’ sub-cluster, as it constitutes one item (see Table 3.4).
Internal consistency reliability coefficients for the Facilitating sub-clusters, indicated in Table
3.4, range from .86 to .96;with the average coefficient being that of .93. The median
Cronbach alpha for the Facilitating sub-clusters is .96; with all reliability coefficients above
.86. The reliability coefficient for the Integrity cluster is .63.
Internal consistency reliability coefficients for the Relationship Harmony sub-clusters range
from .71 to .93. The average reliability coefficient for the sub-clusters is .82, with a median
Cronbach alpha of .82. All internal consistency reliabilities for the sub-clusters of
Relationship Harmony are above .71. Internal consistency reliability coefficients for the
Social Desirability sub-clusters range from .17 to .33, with an average Cronbach alpha of .23.
The median reliability coefficient is .20, with no internal consistency reliability above .33 for
this cluster (see Table 3.4).
Table 3.4also indicates an internal consistency reliability of .79 for the Interpersonal
Relatedness factor of the CPAI-2.Reliability coefficients for the Interpersonal Relatedness
scales range from .36 to .68, with an average coefficient of .55. The median Cronbach alpha
for the Interpersonal Relatedness scales is .59. All reliability coefficients are below .68.
3.6. Construct Validity- SAPI and CPAI-2
This study conducted a factor analysis procedure on the clusters and sub-clusters of the SAPI
and the factor and scales of the CPAI-2, as well as the items of the SAPI, in order to
empirically validate these psychometric instruments by commenting on their construct
validity. All factor loadings above .40 and below -.40 were considered as a loading on
57
thatparticular factor. These cut-off loading points were chosen as a result of their use by
Cheung et al. (2008) and Nel (2008).
The number of factors to be examined for the collective dimensions of the SAPI and CPAI-2
were determined based on theoretical and empirical factors. At a theoretical level the SAPI
postulates four broad collective dimensions- namely Soft-Heartedness, Facilitating, Integrity,
Relationship Harmony- and a fifth factor Social Desirability. These five clusters, in theory,
propose the existence of fourteen sub-clusters that constitute these clusters. The CPAI-2, on
the other hand, postulates one collective dimension- Interpersonal Relatedness- constructed
by six scales; namely Traditionalismvs Modernity, Ren Qing, Social Desirability, Discipline,
Harmony and Thriftvs Extravagance. Empirically, Cattell’s scree plot was examined, as well
as eigenvalues greater than one in order to determine the number of factors suggested.
3.6.1. Factor Analysis- SAPI
3.6.1.1.Scale Level
3.6.1.1.1. Five Factor Solution
A factor analysis with a five factor solution was conducted in order to determine whether the
five clusters as assessed in the SAPI with the suggested sub-clusters do in fact load as five
separate factors. These clusters include Soft-Heartedness, Facilitating, Integrity, Relationship
Harmony and Social Desirability.The eigenvalue Table 3.5, however, suggests the retention
of three overall factors for the SAPI; this is evident by the three numerical values greater than
1.The scree plot,in Figure 3.1, also indicates three factors.The total variance explained by the
five factor solution is 81.04%. Factor 1 explains 40.29% of the variance, Factor 2 explains
20.40%, Factor 3- 9.06%, Factor 4- 6.03% and Factor 5- 5.26%.
58
Table 3.5: Eigenvalues for the SAPI five factor solution
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Eigenvalues of the Correlation Matrix:
Total=14 Average=1
Eigenvalue Difference
Proportion Cumulative
5.64
2.78
0.40
0.40
2.86
1.59
0.20
0.61
1.27
0.42
0.09
0.70
0.84
0.11
0.06
0.76
0.74
0.18
0.05
0.81
0.56
0.09
0.04
0.85
0.48
0.06
0.03
0.88
0.42
0.10
0.03
0.91
0.32
0.06
0.02
0.94
0.25
0.06
0.02
0.95
0.20
0.02
0.01
0.97
0.18
0.03
0.01
0.98
0.15
0.04
0.01
0.99
0.11
0.01
1
Figure 3.1: Cattell’s Scree Plot for SAPI Sub-clusters five factor solution
59
Table 3.6: Five factor solution for the SAPI using varimax rotation
TEncouraging
Others
TGuidance
TActive
Support
THarmony
Maintenance
TEmpathy
THelp Others to
Succeed
THostility
THarmony
Breach
TApproachability
TEgoism
TSocialDesirabilit TSocialDesi
rability
y
TIntegrity
TIntegrity
TImpressionMan
agement
TSelf-Deception
Rotated Factor Pattern
Factor 1
Factor 2
Factor 3
0.88
-0.26
-0.07
Factor 4
0.07
Factor 5
-0.04
0.87
0.85
0.00004
-0.26
-0.01
0.15
-0.002
-0.002
0.04
0.04
0.78
0.45
0.20
0.07
0.002
0.78
0.76
-0.42
0.01
0.23
-0.40
0.10
-0.01
0.007
0.04
-0.02
-0.12
0.89
0.87
0.07
0.003
0.11
0.10
0.08
-0.05
-0.21
-0.17
-0.05
0.82
0.73
0.35
0.18
0.37
0.71
-0.13
-0.05
0.33
0.14
0.09
0.07
0.50
0.08
0.17
0.02
0.60
0.11
-0.12
0.93
0.26
0.16
0.04
0.12
0.12
0.17
0.95
The five factor solution provides evidence of five overall factors, with sub-cluster loadings
above .40 on each of the five factors presented in Table 3.6. Factor 1 includes the sub-clusters
Encouraging Others, Guidance, Active Support, Harmony Maintenance, Empathy, and ‘I help
others to succeed’; with positive loadings between .76 and .88. These sub-clusters are largely
representative of the Soft-Heartedness and Facilitating clusters, with a sub-cluster from the
Relationship Harmony cluster. The sub-clusters Hostility, Harmony Breach, Approachability
and Egoism represent Factor 2, with positive loadings ranging from .73 to .89. These subclusters form a mix of Soft-Heartedness and Relationship Harmony clusters. Factor 3 is
representative of the Integrity cluster and Social Desirability sub-cluster, with positive
loadings of .71 and .60. The sub-cluster Impression Management represents Factor 4 with a
positive loading of .93; while the sub-cluster Self-deception represents Factor 5 with a
positive loading of .95. The sub-clusters that appear to cross load on two factors include,
Harmony Maintenance, Empathy, ‘I help others to succeed’ and the Integrity cluster. These
factors may therefore require further examination.
60
3.6.1.1.2. Three Factor Solution
The eigenvalues in Table 3.5 suggest the retention of three factors, thus a three factor solution
was examined in order to determine whether all the sub-clusters of the SAPI load onto three
factors.The total variance explained by the three factor solution is 69.75%. Factor 1 explains
40.29% of the variance, Factor 2 explains 20.40% and Factor 3- 9.06%.
Table 3.7: Three factor solution for the SAPI using varimax rotation
Rotated Factor Pattern
Factor 1
0.86
TEncouraging
Others
TActiveSupport
TGuidance
THarmony
Maintenance
TEmpathy
THelp Others to Succeed
TIntegrity
TIntegrity
THostility
TApproachability
THarmony
Breach
TEgosim
TImpressionManagement
TSelf-Deception
TSocialDesirability
TSocialDesirability
Factor 2
-0.29
Factor 3
-0.03
0.86
0.86
0.80
-0.24
-0.1
-0.42
0.10
-0.002
0.16
0.80
0.72
0.57
-0.22
-0.20
-0.14
-0.40
-0.06
0.29
0.87
0.85
0.83
0.20
-0.23
0.40
0.14
0.08
0.02
-0.15
0.02
0.06
-0.02
0.79
-0.10
0.15
0.41
0.21
0.73
0.67
0.65
The three factor solution provided in Table 3.7 provides evidence of three overarching SAPI
factors. Factor 1 includes the sub-clusters Encouraging Others, Active Support, Guidance,
Harmony Maintenance, Empathy, ‘I help others to succeed’ and the Integrity cluster; positive
loadings range between .57 and .86. Factor 2 includes the sub-clusters Hostility,
Approachability, Harmony Breach, and Egoism, with positive loadings ranging from .79 to
.87. Factor 1 and Factor 2 thus provide a mix of Soft-Heartedness and Relationship Harmony
sub-clusters; with Factor 1 including the Facilitating and Integrity clusters. The final Factor 3
includes sub-clusters Impression Management, Self-deception and Social Desirabilityrepresentative of the Social Desirability cluster. Cross loadings appear to decrease in the three
61
factor solution, however cross loadings exist for Harmony Maintenance, Integrity and Social
Desirability; indicating that these three sub-clustersmay require further examination.
3.6.1.1.3. Four Factor Solution
A factor analysis with a four factor solution was conducted on the SAPI sub-clusters. This
was done in order to determine whether the sub-clusters indicative of the collective
dimension of the SAPI- Soft-Heartedness, Facilitating, Integrity and Relationship Harmonydo in fact appear as a four factor structure as the theory suggests.The eigenvalue Table 3.8
suggests that two factors should be retained as two eigenvalues have a numerical value
greater than 1.The scree plot in Figure 3.2 also indicates that two factors should be
retained.The total variance explained by the four factor solution is 84.84%. Factor 1 explains
50.66% of the variance, Factor 2 explains 21.57%, Factor 3- 7.60% and Factor 4- 5.01%.
Table 3.8: Eigenvalues for the SAPI four factor solution
1
2
3
4
5
6
7
8
9
10
11
Eigenvalues of the Correlation Matrix:
Total=11 Average=1
Eigenvalue
Difference Proportion Cumulative
5.57
3.20
0.51
0.51
2.37
1.54
0.22
0.72
0.84
0.28
0.08
0.80
0.55
0.13
0.05
0.85
0.42
0.08
0.04
0.89
0.34
0.08
0.03
0.92
0.26
0.06
0.02
0.94
0.20
0.02
0.02
0.96
0.18
0.03
0.02
0.98
0.15
0.04
0.01
0.99
0.11
0.01
1
Figure 3.2: Cattell’s Scree Plot for SAPI Sub-clusters four factor solution
62
Table 3.9: Four factor solution for the SAPI using varimax rotation
Rotated Factor Pattern
Factor 1
Factor 2
0.88
-0.17
0.87
-0.32
0.84
-0.37
0.81
-0.20
0.72
0.01
-0.098
0.93
-0.21
0.91
-0.31
0.73
-0.30
0.59
0.36
-0.11
TActiveSupport
TEmpathy
THarmonyMaintenance
TEncouragingOthers
TGuidance
THarmonyBreach
THostility
TApproachability
TEgoism
THelp Others to
Succeed
TIntegrity
TIntegrity
0.47
0.09
Factor3
0.18
0.08
0.13
0.40
0.48
-0.08
-0.09
-0.03
0.01
0.88
Factor 4
0.08
0.08
0.08
-0.09
0.095
-0.06
0.11
0.399
0.58
-0.002
0.02
0.79
Table 3.9indicates the existence of four factors. The sub-clusters Active Support, Empathy,
Harmony Maintenance, Encouraging Others and Guidance have a strong positiveloading onto
Factor 1; with loadings between .72 and .88. This factor thus represents a mixture of SoftHeartedness, Facilitating and Relationship Harmony sub-clusters. The sub-clustersHarmony
Breach, Hostility, Approachability and Egoism have positive loadings on Factor 2, with
loadings between .59 and .73. Factor 2 therefore includes an amalgamation of SoftHeartedness and Relationship Harmony sub-clusters. Factor 3 includes a positive loading
63
from the sub-cluster‘I help others to succeed’; indicative of the Soft-Heartedness cluster.
Finally, Factor 4 includes a positive loading from the Integrity cluster. The sub-clusters
Encouraging Others, Guidance, Egoism and Integrity appear to be cross loading onto two
factors and may require further examination.
3.6.1.1.4. Two Factor Solution
The eigenvalues for the four factor solution suggeststhe retention of twofactors as indicated
by Table 3.8; thus a two factor solution of the SAPI sub-clusters was examined.The total
variance explained by the two factor solution is 72.23%. Factor 1 explains 50.66% of the
variance and Factor 2 explains 21.57%.
Table 3.10: Two factor solution for the SAPI using varimax rotation
Rotated Factor Pattern
Factor 1
0.87
TActiveSupport
0.85
TEncouragingOthers
0.84
TGuidance
0.83
TEmpathy
0.82
THarmonyMaintenance
0.66
THelp Others to
Succeed
TIntegrity
0.63
TIntegrity
-0.21
THostility
-0.19
TApproachability
-0.11
TEgoism
-0.15
THarmony
Breach
Factor 2
-0.22
-0.31
-0.03
-0.36
-0.395
-0.13
0.37
0.88
0.86
0.82
0.81
The two factor solution provided in Table 3.10 provides evidence of two overarching SAPI
factors. Factor 1 includes the sub-clusters Encouraging Others, Active Support, Guidance,
Harmony Maintenance, Empathy, ‘I help others to succeed’ and the Integrity cluster; positive
loadings range between .63 and .87. Factor 2 includes the sub-clusters Hostility,
Approachability, Harmony Breach, and Egoism, with positive loadings ranging from .81 to
64
.88. Factor 1 and Factor 2 thus provide a mix of Soft-Heartedness and Relationship Harmony
sub-clusters, with Factor 1 including the Facilitating and Integrity clusters.
3.6.1.2. Item Level
Factor analysis of the SAPI items indicates the existence of two factors. Although the
eigenvalues greater-than-one rule suggests seventy four factors; the eigenvalue Table
3.11depicts the clear presence of two factors, as these factors possess eigenvalues much
higher than the rest of the factors. Cattell’s scree plot presented in Figure 3.3; however
indicates the retention of five factors (see Table 4.3 in Chapter 4).
Table 3.11:Eigenvalues for the SAPI at item level
1
2
3
4
5
6
7
Eigenvalues of the Correlation Matrix:
Total=290 Average=1
Eigenvalue Difference
Proportion
Cumulative
55.50
37.25
0.19
0.19
18.25
9.70
0.06
0.25
8.55
1.73
0.03
0.28
6.82
1.63
0.02
0.31
5.19
0.91
0.02
0.33
4.28
0.11
0.01
0.34
4.17
0.24
0.01
0.35
Figure 3.3: Cattell’s Scree Plot for SAPI at item level
65
3.6.1.2.1. Five Factor Solution
A five factor varimax solution was conducted at item level for the SAPI in order to determine
if construct validity is present, by exploring whether the items load onto the five factors
proposed by the SAPI developers. The total variance explained by the five factor solution at
item level is 32.52%. Factor 1 explains 19.14% of the variance, Factor 2 explains 6.29%,
Factor 3- 2.95%, Factor 4- 2.35% and Factor 5- 1.79%.
From the varimax solution it is evident that the five factors do not load as expected, with only
four of the five factors loading with a factor loading of above .40 and below -.40. Factor 1
and Factor 3 are an amalgamation of 31 Soft-Heartedness items and 23 Relationship
Harmony items. The Soft-Heartedness sub-clusters Empathy (21 items) and Active Support
(8 items) load positively onto Factor 1 with moderate to high loadings. The sub-clusters
Hostility (Q41) and Egoism (Q205) of Soft-Heartedness have one moderate positive loading
each on Factor 1. The Relationship Harmony sub-clusters Harmony Maintenance (20 items)
and Approachability (3 items) load positively onto Factor 1, with moderate to high loadings.
One item (Q200) from the Approachability sub-cluster loaded onto Factor 1 negatively with a
loading of -.43. One item (Q235) from the Integrity cluster loaded onto Factor 1 with a
moderate loading of .51.
66
Factor 2 can be described as the Facilitating cluster. The sub-clusters Guidance (41 items)
and Encouraging Others (6 items) have moderate to high positive loadings on Factor 2. This
factor also has a high loading of .70 from Hostility (Q55) of the Soft-Heartedness cluster, a
moderate loading of .41 from the Empathy sub-cluster and a high loading of .60 from the ‘I
help others to succeed’ sub-cluster of this factor. Factor 2 includes 6 moderate to high
loadings each from the Soft-Heartedness sub-cluster Active Support and the Relationship
Harmony sub-cluster Harmony Maintenance.
Factor 3 as stated is an amalgamation of 31 Soft-Heartedness items, 23 Relationship
Harmony items and 7 Integrity items.The Soft-Heartedness sub-clusters Hostility (25 items)
and Egoism (6 items) load positively onto Factor 3 with moderate to high loadings. The
Relationship Harmony sub-clusters Harmony Breach (15 items) and Approachability (8
items) load positively onto Factor 3 with moderate to high loadings. 7 items for the Integrity
cluster load positively with a moderate to high loading on Factor 3. 1 item (Q14) from the
Impression Management sub-cluster loaded onto Factor 3 negatively with a loading of -.47. 1
item (Q54) from the Guidance sub-cluster of the Facilitating cluster loaded onto Factor 3 with
a moderate loading of .54. 4 items from the Social Desirability cluster indicated positive
moderate loadings on Factor 3; namely from the Social Desirability sub-cluster (3 items) and
the Impression Management sub-cluster (1 item).
Factor 4 of the varimax solution is a combination of 4 Integrity items and 4 Social
Desirability items loading positively with moderate to high loadings. The Integrity items
loaded mostly moderately with one high loading that range from .41 to .58. The Social
Desirability items included 2 high loadings from the Social Desirability sub-cluster and 2
moderate loadings from the Self-deception sub-cluster. Factor 5 obtained no loadings above
.40 or below -.40.
The 5 factor varimax solution presents a total of 96 of 290 items that do not load onto any of
the five suggested factors. Of these 96 items with no loadings; 29 are from the SoftHeartedness cluster, 4 from the Facilitating cluster, 22 from the Integrity cluster, 24 from the
Relationship Harmony cluster and 17 from the Social Desirability cluster. From the Soft-
67
Heartedness cluster- 7 Active Support, 10 Empathy, 3 Egoism and 9 Hostility items did not
load anywhere. From the Relationship Harmony cluster- 18 Harmony Maintenance, 2
Harmony Breach and 4 Approachability items did not load on any factor. 3 Guidance items
and 1 Encouraging Others item from the Facilitating cluster did not load anywhere. Finally,
from the Social Desirability cluster, 7 Social Desirability, 5 Impression Management and 5
Self-deception items did not load on any one of the five factors. These items thus needed to
be reworked or removed from the SAPI.
3.6.1.2.2. Fourteen Factor Solution
A fourteen factor varimax solution was conducted at item level for the SAPI in order to
determine if construct validity is present, by exploring whether the items load onto the
fourteen sub-clusters proposed by the SAPI developers. The total variance explained by the
14 factor solution is 43.73%.Factor 1 explains 19.14% of the variance, Factor 2 explains
6.29%, Factor 3- 2.95%, Factor 4- 2.35%, Factor 5- 1.79%, Factor 6- 1.48%, Factor 71.44%, Factor 8- 1.36%, Factor 9- 1.31%, Factor 10- 1.26%, Factor 11- 1.21%, Factor 121.08%, Factor 13- 1.06% and Factor 14- 1.01%.
From the varimax solution it is evident that the fourteen factors do not load as expected, with
only thirteen of the fourteen factors obtaining factor loadings of above .40 and below -.40.
Factor 1 is largely comprised of moderate to high positive loadings of Guidance items (35
items) from the Facilitating cluster. The sub-cluster Encouraging Others (6 items) of the
Facilitating cluster has positive moderate loadings on Factor 1. Four Soft Heartedness subclusters obtained moderate to high positive loadings on Factor 1. 7 items from Active Support
obtained loadings, 1 Hostility itemloaded with a loading of .73, the ‘I help others to
succeed’sub-cluster (1 item- Q275) had a loading of .62 and 1 Empathy item (Q240) obtained
a loading of .43. The sub-cluster Harmony Maintenance (8 items) of the Relationship
Harmony cluster loaded with positive moderate to high loadings ranging from between .40
and .62.
68
Factor 2 of the fourteen factor solution incorporates an amalgamation of Soft-Heartedness,
Relationship Harmony, Facilitating, Integrity and Social Desirability items. This factor
constitutes mostly of Hostility and Harmony Breach items loading positively from moderate
to high. 5 Integrity and 7 Approachability items loaded positively with moderate to high
loadings on Factor 2. 2 Social Desirability and 2 Egoism items loaded positively with
moderate to high loadings on this factor. 1 Guidance item (Q54) loaded with a positive
loading of .59 and 1 Impression Management item (Q14) loaded negatively onto Factor 2
with a loading of -.45.
Factor 3 is mainly comprised of Empathy (16 items) and Harmony Maintenance items (15
items) with positive moderate to high factor loadings. Factor 3 also includes moderate to high
loadings from the Soft-Heartedness sub-clusters Active Support (5 items), Hostility (1 itemQ41) and Egoism (1 item- Q205). 2 items from the Encouraging Others sub-cluster loaded
with moderate to high loadings and 2 items from the Approachability sub-cluster loaded with
moderate loadings. 1 Integrity item (Q235) loaded positively with a factor loading of .48 onto
Factor 3.
Factor 4 incorporates 6 moderate factor loadings from the Integrity, Guidance, Harmony
Maintenance and Active Support sub-clusters. 3 Integrity items, namely Q212, Q195 and
Q122 loaded positively onto Factor 3, as well as an item from Harmony Maintenance (Q231),
Guidance (Q226) and Active Support (Q164). Factor 5 includes 4 positive moderate factor
loadings; encompassing 2 Egoism loadings (Q125 and Q178) of .50, 1 Social Desirability
loading (Q196) of .54 and one Integrity loading (Q117) of .48.
Factor 6 includes 9 moderate to high factor loadings- 5 positive factor loadings from
Encouraging Others (1 item- Q274), Empathy (3 items- Q250, Q265 and Q273) and
Harmony Maintenance (1 item- Q253), and 4 negative factor loadings from Egoism (2 itemsQ263 and Q264), Social Desirability (1 item- Q252) and Integrity (1 item- Q251). Factor 7
comprises of 5 positive and negative factor loadings- 2 positive Guidance loadings of .56 and
.45, 2 negative moderate Integrity loadings and 1 negative Hostility loading (Q118) of .47.
Factor 8 includes 3 positive moderate loadings of .51, .50 and .43; from Impression
69
Management (2 items- Q69 and Q206) and Integrity (1 item- Q222). 3 Integrity items loaded
negatively with moderate to high loadings onto Factor 8. Factor 9 includes one moderate
positively loading item- Q70- from the Approachability sub-cluster of Relationship Harmony
of .50.
Factor 10 encompasses 2 moderate positive loadings of .47 and .46 of the Harmony
Maintenance (Q22) and Harmony Breach (Q130) sub-clusters respectively. 1 item each from
the sub-clusters Impression Management (Q185), Hostility (Q7) and Egoism (36) make up
Factor 11 with loadings of .52, .50 and .42 respectively. Factor 12 includes 1 loading (Q93)
of .44 from the Guidance sub-cluster of the Facilitating cluster. Factor 13 incorporates 2
moderate positive loadings (Q4 and Q193) from the Harmony Maintenance sub-cluster of the
Relationship Harmony cluster of .49 and .44.
The fourteen factor varimax solution presents a total of 90 of 290 items that do not load onto
any of the fourteen suggested factors. Of these 90 items with no loadings; 29 are from the
Soft-Heartedness cluster, 6 from the Facilitating cluster, 17 from the Integrity cluster, 21
from the Relationship Harmony cluster and 17 from the Social Desirability cluster. From the
Soft-Heartedness cluster, 8 Active Support, 12 Empathy, 2 Egoism and 7 Hostility items did
not load anywhere. From the Relationship Harmony cluster 16 Harmony Maintenance and 5
Approachability items did not load on any factor. 5 Guidance items and 1 Encouraging
Others item from the Facilitating cluster did not load anywhere. Finally, from the Social
Desirability cluster, 7 Social Desirability, 2 Impression Management and 7 Self-deception
items did not load on any one of the fourteen factors. These items thus needed to be reworked
or removed from the SAPI.
3.6.2. Factor Analysis- CPAI-2 Scale level
3.6.2.1. Six Factor Solution
A six factor solution was conducted on the CPAI-2 Interpersonal Relatedness collective
dimension, as this factor is believed to be comprised of sixscales that includeTraditionalism
vs Modernity, Ren Qing, Social Desirability, Discipline, Harmony and Thrift vs
70
Extravagance. The eigenvalues presented in Table 3.12 indicate the existence of two factors,
as two eigenvalues are greater than 1.The scree plot presented in Figure 3.4 indicates the
existence of six factors.The total variance explained by the six factor solution is 100%. Factor
1 explains 36.65% of the variance, Factor 2 explains 22.14%, Factor 3- 14.46%, Factor 410.20%, Factor 5- 8.95% and Factor 6- 7.59%.
Table 3.12:Eigenvalues for the CPAI-2 scales
1
2
3
4
5
6
Eigenvalues of the Correlation Matrix:
Total=6 Average=1
Eigenvalue
Difference
Proportion
2.20
0.87
0.37
1.33
0.46
0.22
0.87
0.26
0.14
0.61
0.08
0.10
0.54
0.08
0.09
0.46
0.08
Figure 3.4: Cattell’s Scree Plot for CPAI-2 Scales
Cumulative
0.37
0.59
0.73
0.83
0.92
1
71
Table 3.13: Six factor solution for the CPAI-2 using varimax rotation
THarmony
TTraditionalismvs
Modernity
TThriftvs
Extravagance
TRenQing
TDiscipline
TSocialSensitivity
Factor 1
0.95
-0.03
Rotated Factor Pattern
Factor 2 Factor 3 Factor 4
-0.03
0.07
0.13
0.97
0.02
0.12
Factor 5
0.06
0.19
Factor 6
0.25
-0.04
0.07
0.02
0.98
0.08
0.12
0.09
0.13
0.06
0.26
0.13
0.20
-0.04
0.09
0.13
0.20
0.95
0.15
0.18
0.15
0.95
0.09
0.17
0.08
0.94
The six factor solution presented in table 3.13 indicates the clear existence of six factors with
high positive loadings on each separate factor. Factor 1 consists of the scale Harmony, with a
positive loading of .95. Factor 2 is represented by the scale Traditionalism vs Modernity with
a positive loading of .97. Factor 3 includes the scale Thrift vs Extravagance with a positive
loading of .98. Factor 4 consists of the scale Ren Qing with a positive loading of .95; while
Factor 5 includes the scale Discipline with a positive loading of .95. Finally, Factor 6
incorporates the scale Social Sensitivity with a positive loading of .94.
3.6.2.2. Two Factor Solution
The eigenvalue Table 3.12 indicates the existence of two factors; thus a two factor solution
was conducted on the CPAI-2 scales of the Interpersonal Relatedness factor.The total
variance explained by the two factor solution is 58.78%; with Factor 1 explaining 36.65% of
the variance and Factor 2 explaining 22.14% of the variance.
72
Table 3.14: Two factor solution for the CPAI-2 using varimax rotation
Rotated Factor Pattern
Factor 1
Factor 2
0.84
0.05
TSocialSensitivity
0.82
-0.02
THarmony
0.38
0.35
TThriftvs
Extravagance
-0.20
0.82
TTraditionalismvs
Modernity
0.21
0.77
TDiscipline
0.51
0.52
TRenQing
The rotated factor pattern indicates that the scales Social Sensitivity and Harmony load
positively onto Factor 1 with loadings of .85 and .82 respectively. Factor 2, on the other hand,
consists of the scales Traditionalismvs Modernity, Discipline and Ren Qing with positive
loadings ranging between .82 and .52. The Thrift vs Extravagance scale does not load onto
either of the factors, however its highest loading of .38 is on Factor 1. Further examination on
this factor is therefore required. The scale Ren Qing cross loaded onto both factors indicating
that further examination of this factor may be necessary.
3.6.3. Factor Analysis SAPI and CPAI-2
3.6.3.1. Subscales 1
The SAPI collective dimension sub-clusters and the CPAI-2 Interpersonal Relatedness scales
were placed into a factor analysis in order to determine whether the sub-clusters and scales
load in congruence with the original factor structure.The eigenvalue Table 3.15 indicates four
overall factors to be retained, as they have obtained an eigenvalue greater than 1.The scree
plot presented in Figure 3.5also indicates the existence of four factors.
1
The subscales of the SAPI are referred to as sub-clusters by Nel (2008); while the subscales of the CPAI-2 are
referred to as scales by Cheung et al. (2008). However, when the SAPI and CPAI-2 sub-clusters and scales are
discussed together in this study, they will be referred to in the general term subscale.
73
Table 3.15:Eigenvalues for the SAPI and CPAI-2 subscales
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Eigenvalues of the Correlation Matrix:
Total=20 Average=1
Eigenvalue
Difference
Proportion
6.95
3.83
0.35
3.12
1.72
0.16
1.41
0.07
0.07
1.33
0.35
0.07
0.98
0.13
0.05
0.86
0.11
0.04
0.75
0.03
0.04
0.72
0.15
0.04
0.57
0.05
0.03
0.52
0.05
0.03
0.46
0.03
0.02
0.44
0.04
0.02
0.39
0.05
0.02
0.35
0.06
0.02
0.28
0.03
0.01
0.25
0.05
0.01
0.19
0.02
0.01
0.17
0.03
0.01
0.14
0.04
0.01
0.11
0.01
Cumulative
0.35
0.50
0.57
0.64
0.69
0.73
0.77
0.81
0.83
0.86
0.88
0.91
0.93
0.94
0.96
0.97
0.98
0.99
0.99
1
Figure 3.5: Cattell’s Scree Plot for SAPI and CPAI-2 subscales
74
Table 3.16: Factor solution for the SAPI and CPAI-2 subscales using varimax rotation
Rotated Factor Pattern
Factor 1
Factor 2
0.87
-0.23
0.84
-0.27
TActiveSupport
TEncouragingOt
hers
TGuidance
TEmpathy
THarmonyMaint
enance
THelp Others to
Succeed
TSocialSensitivity
TIntegrity
TIntegrity
THostility
TApproachability
THarmonyBreac
h
TEgoism
THarmony
TSocialDesirabilit TSocialDesi
rability
y
TSelf-Deception
TImpressionMan
agement
TTraditionalismv
s Modernity
TDiscipline
TRenQing
TThriftvs
Extravagance
Factor 3
0.08
-0.08
Factor 4
0.05
0.14
0.82
0.81
0.80
0.01
-0.39
-0.42
-0.06
0.18
0.13
0.25
0.02
0.08
0.67
-0.04
-0.29
0.24
0.63
0.57
-0.21
-0.20
-0.11
-0.38
0.30
0.86
0.84
0.81
0.10
0.40
0.15
0.10
0.06
-0.02
0.07
-0.09
-0.01
-0.14
-0.16
0.46
-0.01
0.77
-0.58
0.39
0.24
0.20
0.67
0.07
-0.04
0.03
0.05
0.02
0.14
-0.07
0.65
0.65
0.13
0.05
-0.09
-0.03
-0.06
0.86
0.32
0.34
0.27
0.08
-0.25
-0.07
0.19
0.27
0.14
0.70
0.47
0.30
The rotated factor pattern in Table 3.16 suggests a four factor solution for the sub-clusters of
the SAPI and scales of the CPAI-2 collective dimension. The sub-clusters Active Support,
Encouraging Others, Guidance, Empathy, Harmony Maintenance, ‘I help others to
succeed’and Integrity of the SAPI, and the Social Sensitivity scale of the CPAI-2 load onto
Factor 1 with positive loadings ranging from .57 to 87. This factor indicates sub-clusters
representative of the SAPI Soft-Heartedness, Integrity, Guidance and Relationship Harmony
clusters and Social Sensitivity from the CPAI-2 Interpersonal Relatedness factor. Factor 2 has
loadings of between -.58 and .86 from the sub-clusters Hostility, Approachability, Harmony
75
Breach, Egoism and the Harmony scale. These sub-clusters are representative of the SoftHeartedness and Relationship Harmony sub-clusters of the SAPI, as well as the Harmony
scale of the CPAI-2 Interpersonal Relatedness factor. All sub-clusters load positively onto
Factor 2; however the scale Harmony loads negatively onto this factor as expected. Factor 3
includes the sub-clusters Social Desirability, Self-deception and Impression Management of
the SAPISocial Desirability cluster; with positive loadings between .67 and .65. Factor 4
includes the scales Traditionalismvs Modernity, Discipline and Ren Qing of the CPAI-2
Interpersonal Relatedness factor, with positive factor loadings of .47 to .86. The Thrift vs
Extravagance scale from the CPAI-2 Interpersonal Relatedness factor did not load onto any
of the four factors; however its highest loading of .30 was on Factor 4. Further examination
of this scale may be required.
3.6.3.2. Scales 2
The SAPI collective dimension clusters- Soft-Heartedness, Facilitating, Integrity and
Relationship Harmony- and the CPAI-2 Interpersonal Relatedness factor were placed into a
factor analysis in order to determine whether they load in congruence with the original
proposed factor structure.The eigenvalue Table 3.17 indicates the retention of two factors,
evidenced by the two numerical values greater than 1.The scree plot presented in Figure 3.6
indicates the existence of six factors.
Table 3.17:Eigenvalues for the SAPI and CPAI-2 scales
1
2
3
4
5
6
2
Eigenvalues of the Correlation Matrix:
Total=6 Average=1
Eigenvalue Difference Proportion Cumulative
3.04
1.90
0.51
0.51
1.14
0.40
0.19
0.697
0.74
0.21
0.12
0.82
0.53
0.21
0.09
0.91
0.32
0.09
0.05
0.96
0.23
0.04
1
The scales of the SAPI are referred to as clusters by Nel (2008); while the scales of the CPAI-2 are referred to
as factors by Cheung et al. (2008). However, when the SAPI and CPAI-2 clusters and factors are discussed
together in this study, they will be referred to in the general term scale.
76
Figure 3.6: Cattell’s Scree Plot for SAPI and CPAI-2 Scales
Table 3.18: Factor solution for the SAPI and CPAI-2 scales using varimax rotation
Rotated Factor Pattern
Factor 1
0.79
TSocialDesirability
0.78
TSoft-Heartedness
0.70
TIntegrity
0.66
TReltionshipHarmony
0.21
TFacilitating
0.05
TInterpersonalRelatedness
Factor 2
-0.21
0.38
0.32
0.55
0.87
0.79
The rotated factor patter presented in Table 3.18 indicates the existence of two overarching
factors. The clusters Social Desirability, Soft-Heartedness, Integrity and Relationship
Harmony of the SAPI collective dimension load positively onto Factor 1, with loadings
between .66 and .79. Factor 2 includes the SAPI cluster Facilitating and the CPAI-2
Interpersonal Relatedness factor with positive factor loadings of .87 and .79 respectively. The
cluster Relationship Harmony of the SAPI cross loaded onto both these factors, indicating the
possible need for further examination.
3.6.3.3. Six Factor Solution
77
The SAPI collective dimension clusters- Soft-Heartedness, Facilitating, Integrity and
Relationship Harmony- and the CPAI-2 Interpersonal Relatedness factor were placed into a
six factor solution in order to determine whether these dimensions load in congruence with
the original factor structure proposed by the theory.
Table 3.19: Six factor solution for the SAPI and CPAI-2 using varimax rotation
TInterpersonalRelatedness
TSocialDesirability
TIntegrity
TFacilitating
TRelationshipHarmony
TSoft-Heartedness
Rotated Factor Pattern
Factor Factor 2 Factor
1
3
0.96
0.05
0.11
0.05
0.97
0.15
0.12
0.17
0.93
0.30
-0.03
0.14
0.14
0.12
0.24
0.06
0.22
0.25
Factor
4
0.24
-0.02
0.13
0.89
0.25
0.23
Factor
5
0.11
0.09
0.20
0.23
0.85
0.35
Factor
6
0.05
0.15
0.20
0.20
0.34
0.84
The rotated factor pattern in Table 3.19 indicates the existence of six overarching factors of
the SAPI and CAPI-2 collective dimensions. Interpersonal Relatedness of the CPAI-2 loaded
positively onto Factor 1 with a factor loading of .96. Social Desirability of the SAPI loaded
positively onto Factor 2 with a factor loading of .97. Integrity of the SAPI loaded onto Factor
3 with a positive factor loading of .93. Facilitating of the SAPI loaded onto Factor 4 with a
positive factor loading of .89. Relationship Harmony of the SAPI loaded positively onto
Factor 5 with a factor loading of .85 and finally Soft-Heartedness of the SAPI loaded onto
Factor 6 with a positive factor loading of .84.
3.7. Construct Bias for the SAPI and CPAI-2
Construct bias was examined at the scale level of the two instruments- the SAPI and the
CPAI-2- using ANOVAs.Evidence for construct bias was examined across three variables;
78
namely gender, race and home language. Parametric one-way ANOVAs were used for all
comparisons provided that the conditions of normality and homogeneity of variance were
met. In cases where one or both of the conditions above were not met, the non-parametric
Kruskal-Wallis one-way ANOVA technique was used. Parametric and non-parametric results
are presented in tables 3.26, 3.27 and 3.28 below; however the non-parametric results have
been italicised. Mean scores were examined for significant parametric results and mean rank
scores were examined for non-parametric results. The effect sizes of the significant results of
the parametric ANOVAs were examined using Cohen’s d.
Results for construct bias are presented below for gender, followed by race and then home
language.
3.7.1. Gender- The SAPI and CPAI-2
Table 3.26 presents the ANOVA results obtained for gender and the scales and subscales of
the SAPI and CPAI-2. All the scales and subscales were normally distributed, except for the
clusterSoft-Heartedness and the sub-clusterActive Supportof the SAPI and the Social
Sensitivityscale of the CPAI-2.The homogeneity of variance criterion was also not met for the
Integrity and Relationship Harmony clusters of the SAPI. Hence non-parametric KruskalWallis ANOVAs were conducted for these scales and subscales.
From Table3.26 it is evident that significant differences occur between gender and
Interpersonal Relatedness, Ren Qing, Social Sensitivity, Discipline, Harmony, Egoism,
Hostility, Empathy, Approachability and Harmony Breach at the .01 level of significance. At
the .05 level of significance, significant differences were found between gender and Active
Support, and Harmony Maintenance. There are no other statistically significant scores for the
scales and subscales of the SAPI and CPAI-2.
Table 3.20: ANOVA results for gender on the SAPI and CPAI-2
Scale
Interpersonal
Df
F/X2
p
1,475
17.22
<.0001**
Mean Scores/Ranks
M
F
235.16
243.68
Cohen’s d
0.46
79
Relatedness
Traditionalismvs
Modernity
Ren Qing
Social
Sensitivity
Discipline
Harmony
Thriftvs
Extravagance
Soft
Heartedness
Egoism
Hostility
Empathy
Active Support
I help others to
succeed
Facilitating
Guidance
Encouraging
Others
Integrity
Integrity
Relationship
Harmony
Approachability
Harmony
Maintenance
Harmony
Breach
Social
Desirability
Social
Desirability
Impression
Management
Self-deception
*p < 0.05
1,475
0.42
0.5197
1,475
1
16.57
15,0535
<.0001**
0.0001**
40.375
192.74
42.50
251.90
0.45
1,475
1,475
1,475
10.28
9.18
0.28
0.0014**
0.0026**
0.5988
33.61
49.12
35.39
51.10
0.36
0.34
1
2.8358
0.0922
1,475
1,475
1,475
1
1,472
9.88
27.23
12,79
5.6691
0.05
0.0018**
<.0001**
0.0004**
0.0173*
0.8277
27.36
94.27
123.10
210.56
25.61
84.59
128.34
246.93
0.35
0.58
0.40
1,475
1,475
1,475
0.1
0.01
3.43
0.7574
0.9397
0.0648
1
1
1
0.7395
0.7395
0.5350
0.3898
0.3898
0.4645
1,475
1,475
21.06
4.69
<.0001**
0.0308*
39.68
168.26
36.44
172.44
0.51
0.24
1,475
10.65
0.0012**
42.30
39.27
0.36
1,475
1.23
0.2684
1,475
0.39
0.5303
1,475
0.05
0.818
1,475
2.43
0.12
**p < 0.01
An examination of the mean scores and mean rank scores for significant differences between
gender and the scales and subscales of the SAPI and CPAI-2 reveals that in this sample,
females scored higher than males on Interpersonal Relatedness, Ren Qing, Social Sensitivity,
80
Discipline, Harmony, Empathy, Active Support and Harmony Maintenance. Females,
however, score lower than males onEgoism, Hostility, Approachability and Harmony Breach.
Moderate effect sizes between .34 and .46 were obtained for differences on the Interpersonal
Relatedness factor, and the RenQing, Discipline and Harmony scales of the CPAI-2; as well
as the SAPI sub-clusters Egoism, Empathy and Harmony Breach. The sub-cluster Harmony
Maintenance of the SAPIdemonstrates a small effect size of .24; while the SAPI sub-clusters
Hostility (d = .58) and Approachability (d = .51) reveal large effect sizes.
3.7.2. Race- The SAPI and CPAI-2
Table 3.21represents the results obtained for race and the scales and subscales of the SAPI
and CPAI-2. All the scales and subscales were normally distributed except forthe SoftHeartednesscluster and the Active Supportsub-cluster of the SAPI and the Social
Sensitivityscale of the CPAI-2. The homogeneity of variance criterion was also not met for
the Interpersonal Relatedness, Discipline and Thriftvs Extravagancescales of the CPAI-2.
Hence non-parametric Kruskal-Wallis ANOVAs were conducted for these scales.
From Table 3.21it is evident that significant differences occur between race and
Interpersonal Relatedness, Traditionalismvs Modernity, Egoism, Hostility, Facilitating,
Guidance, Encouraging Others, Integrity, Approachability, Harmony Breach and Impression
Management at the .01 level of significance. Significant differences occur between race and
Harmony Maintenance and ‘I help others to succeed’ at the .05 level of significance. Aside
from those reported; there were no other significant differences between race and the scales
and subscales of the SAPI and CPAI-2.
Table 3.21: ANOVA results for race on the SAPI and CPAI-2
Scale
Interpersonal
Df
F/X2
P
1
14,2290
0.0002**
Mean Scores/Ranks
W
NW
203,47
250,40
Cohen’s
D
81
Relatedness
Traditionalismvs
Modernity
Ren Qing
Social
Sensitivity
Discipline
Harmony
Thriftvs
Extravagance
Soft
Heartedness
Egoism
Hostility
Empathy
Active Support
I help others to
succeed
Facilitating
Guidance
Encouraging
Others
Integrity
Integrity
Relationship
Harmony
Approachability
Harmony
Maintenance
Harmony
Breach
Social
Desirability
Social
Desirability
Impression
Management
Self-deception
*p < 0.05
1,456
70.32
<.0001**
36.91
41.94
0.79
1,456
1
3.05
0.32
0.0815
0.5699
1
1,456
1
1.46
0.59
0.11
0.2276
0.4422
0.7457
1
2.56
0.1093
1,456
1,456
1,456
1
1, 454
8.73
7.44
1.15
0.47
5.25
0.0033**
0.0066**
0.2840
0.4934
0.0224*
26.70
88.99
25.31
84.59
0.28
0.26
3.60
3.77
0.22
1,456
1,456
1,456
16.54
16,96
8.61
<.0001**
<.0001**
0.0035**
196.01
157.28
38.73
205.64
165.48
40.16
0.38
0.39
0.28
1,456
1,456
1,456
15.14
15.14
2.10
0.0001**
0.0001**
0.1479
116.74
116.74
113.86
113.86
0.37
0.37
1,456
1,456
19.11
4,97
<.0001**
0.0263*
38.51
169.68
35.89
173.30
0.41
0.21
1,456
15.97
<.0001**
41.63
38.48
0.38
1,456
3.06
0.0811
1,456
0.16
0.6858
1,456
23.00
<.0001**
20.54
21.78
0.45
1,456
0.17
0.6799
**p < 0.01
An examination of the mean scores and mean rank scores for significant differences between
race and the scales and subscales of the SAPI and CPAI-2 reveals that in this sample, nonWhites scored higher than Whites on Interpersonal Relatedness, Traditionalismvs Modernity,
82
Facilitating, Guidance, Encouraging Others, ‘I help others to succeed’, Harmony
Maintenance and Impression Management. Non-Whites, however, scored lower than Whites
on Egoism, Hostility, Integrity, Approachability and Harmony Breach.
Small effect sizes were obtained for differences on the SAPI sub-clusters Egoism (d = .28), ‘I
help others to succeed’ (d = .22), Hostility (d = .26), Encouraging Others (d = .28) and
Harmony Maintenance (d = .21). Moderate effect sizes ranging between .37 and .45 were
obtained for differences demonstrated by the SAPI clusters Facilitating and Integrity, as well
as the SAPI sub-clusters Guidance, Integrity, Approachability, Harmony Breach and
Impression Management. A very large effect size was obtained for the Traditionalism vs
Modernity scale of the CPAI-2, d = .79.
3.7.3. HomeLanguage- The SAPI and CPAI-2
Table 3.22represents the results obtained for home language and the scales and subscales of
the SAPI and CPAI-2. All the scales and subscales were normally distributed except forSoftHeartednesscluster and Active Supportsub-cluster of the SAPI and the Social Sensitivityscale
of the CPAI-2. The homogeneity of variance criterion was also not met for theRen Qing scale
of the CPAI-2 and the Hostility sub-cluster of the SAPI. Hence non-parametric KruskalWallis ANOVAs were conducted on this scale and sub-cluster.
From Table 3.22 it is evident that significant differences occur between home language and
Interpersonal Relatedness, Traditionalismvs Modernity, ‘I help other to succeed’, Facilitating,
Guidance, Encouraging Others, Integrity, Approachability and Impression Management at the
.01 level of significance. Aside from those reported, there were no other statistically
significant differences between home language and the scales and subscales of the SAPI and
CPAI-2.
Table 3.22: ANOVA results for home language on the SAPI and CPAI-2
Scale
Df
F/X2
p
Mean Scores/Ranks
E1
E2
Cohen’s
D
83
Interpersonal
Relatedness
Traditionalismvs
Modernity
Ren Qing
Social
Sensitivity
Discipline
Harmony
Thriftvs
Extravagance
Soft
Heartedness
Egoism
Hostility
Empathy
Active Support
I help others to
succeed
Facilitating
Guidance
Encouraging
Others
Integrity
Integrity
Relationship
Harmony
Approachability
Harmony
Maintenance
Harmony
Breach
Social
Desirability
Social
Desirability
Impression
Management
Self-deception
*p < 0.05
1,466
9.49
0.0022**
239.82
245.34
0.30
1,466
25.39
<.0001**
38.40
41.66
0.48
1
1
0.56
0.16
0.5429
0.6932
1,466
1,466
1,466
2.32
0.08
3.50
0.1288
0.7762
0.0622
1
0.01
0.9181
1,466
1
1,466
1
1,463
3.82
0.41
0.98
0.32
8.11
0.0511
0.5209
0.3223
0.5740
0.0046**
3.60
3.83
0.27
1,466
1,466
1,466
30.88
33.20
12.08
<.0001**
<.0001**
0.0006**
195.97
157.18
38.80
209.27
168.75
40.53
0.53
0.55
0.33
1,466
1,466
1,466
8.91
8.91
0.13
0.0030**
0.0030**
0.7151
116.03
116.03
113.75
113.75
0.29
0.29
1,466
1,466
7.27
2.49
0.0073**
0.1152
37.81
36.12
0.26
1,466
3.48
0.0626
1,466
0.80
0.3724
1,466
0.46
0.4967
1,466
8.65
0.0034**
20.93
21.72
0.28
1,466
0.14
0.7111
**p < 0.01
An examination of the mean scores and mean rank scores for significant differences between
home language and the scales and subscales of the SAPI and CPAI-2 reveals that in this
84
sample, English second language speakersscored higher than English first language speakers
on Interpersonal Relatedness, Traditionalismvs Modernity, ‘I help others to succeed’,
Facilitating, Guidance, Encouraging Others and Impression Management. English second
language speakers, however, scored lower than English first language speakers on Integrity
and Approachability.
Small effect sizes of .29, .26 and .28 were demonstrated for differences on Integrity,
Approachability and Impression Management of the SAPI respectively. A small to moderate
effect size of .30 was obtained for the Interpersonal Relatedness factor of the CPAI-2.
Moderate effect sizes of .48 were obtained for the Traditionalismvs Modernityscale of the
CPAI-2 and .33 for the SAPI Encouraging Others sub-cluster. Large effect sizes were
obtained for The SAPI cluster Facilitating (d = .53) and the SAPI sub-cluster Guidance (d =
.55).
3.8. Qualitative Data: Thematic Content Analysis
Nineteen volunteers participated in two focus groups for this study. The responses from both
of the focus groups conducted will be examined together in this section. This will be done as
both of the focus groups appeared to share similar dynamics in terms of the issues identified,
mood, tone and atmosphere at the time that they were conducted. Instances in which themes
discussed were only addressed by one of the two focus groups were highlighted and detailed.
The data was analysed using thematic content analysis and the following seven themes
emerged; namely a general understanding of personality, the innate existence of personality
vs the overt expression of personality, collectivism in personality, a national identity, culture,
psychometric testing and social desirability. The most dominant themes to emerge by all the
participants of the focus group was that of the innate existence of personality vs the overt
expression of personality, collectivism in personality, a national identity and culture. These
themes will be discussed below using pseudonyms such as J1 and F5 in order to preserve the
anonymity and confidentiality of the participants.
3.8.1. General Understandings of Personality
85
The term personality commands a wide array of understandings throughout the world and
particularly within psychology. Thus a number of definitions and conceptualisations can be
provided to illustrate a general understanding of the core of personality from an individual
perspective. Some of these conceptualisations are provided by focus group members and will
be discussed below.
Personality is described by participant F1 to be “Peoples’ characteristics” or “the portrayal of
peoples’ individualities,” according to participant F2.Participant J1 continued establishing
that “personality is something specific to a person and it is individual… I think people try to
categorise personality [for example] into type A and type B, but I think there aren’t just two;
there are many personalities.” Participant J3 elaborates explaining, “there are a lot of things
that influence who you are and the way you define yourself and your personality.”
Participants of focus group 2 (n=3) exemplified the difference between personality and
character, stating “I have always thought that personality you can change, but character you
can’t” (Participant F3). According to participant F3 character is “the essence of someone”,
while personality “is more social constructs that you kind of adapt to… that becomes your
way of viewing the world and your association.” Thus personality, as defined by these
participants (n=3), is how you view yourself and how you present yourself to the world.
Participant F2 demonstrated this by stating, “personality is who you are and how that comes
out to everyone else. Character is your fundamentals.” Participant F1 explained personality to
be a broad concept in which no two people possess the same personality and so as a result it
cannot be categorised. This discussion signifies that an understanding of personality cannot
be categorised or boxed, thus indicating a move towards an understanding of traits rather than
types of personalities.
3.8.2. The Innate Existence of Personality vs the Overt Expression of
Personality
86
Personality, as defined by focus group participants, is depicted to be a result of an
amalgamation between one’s innate personality that is presented at birth and the overt
expression of this personality as shaped by one’s external circumstances. It is within these
confines that one’s personality is said to exist by these focus group members.
All participants (n=19) within the two focus groups highlighted the paradox evident within
the innate existence and overt expression of personality. Participant F1 began this discussion
establishing that personality “is who you are when you are alone, it is what makes you
completely… but it can change if you are alone in a crowd.” This brings to light both aspects
of the paradox present within personality. Participant F2 explained that personality is
something you are born with and is moulded by your experiences and the people around you,
“some people are just a product of where they come from [overt], whereas other people are
just born that way [innate]”. Thus this participant (F2) reasoned that people “have certain
predispositions, but it is brought out in the environment”. This indicates that the 19
participants consider personality to be a combination of innate personality variables
explicated from birth, as well as variables contracted by one’s overt environment.
However some participants (n=4) recognised one’s innate personality, but emphasised the
effect of one’s environment on their personality. Participant J2 stressed that “a person’s
personality changes”; accordingly environmental factors influence the overt expression of
personality and would consequently cause the same person to be categorised as possessing
different traits over time. Participant F2 agreed that “It depends on the environment that you
are in” however stating that “[personality] can change, but it won’t change easily.”
Participant F4 argued, in contrast, that we are rather able to adapt the personality traits that
we overtly reveal to others in specific situations than change our personality fundamentally as
a result of the environment. Participant J2 continued, “I think personality is a learnt thing. I
think it is adapted.” This participant (J2) indicated that one’s background, upbringing and
history are all contributing factors to one’s personality. Participant J3 elaborates on this point
arguing that “you cannot ignore the influence of past experiences, as well as environmental
factors; like where you grew up or the cultural factors that affect your personality.”
Participant J2 believed that our community helps build and mould who we are, “we will
constantly influence each other and I think whether you believe in the concept of the spirit of
87
the community, I think indirectly we are the spirit of the community in terms of the way we
influence each other all the time.” These participant (n=4) thus emphasised one’s
environmental influences as a fundamental determinant of personality.
3.8.3. Collectivism in Personality
The individualism/collectivism debate within non-Western and Western cultures was
discussed in the focus groups, with particular reference to the emergence of these opposing
personality traits within personality in South Africa.
Participants (n=19) in the focus group felt that personality in South Africa is a merger
between different cultural influences. Participants (n=2) described it to be a fusion of
collectivistic and individualistic traits, as a result of Western and non-Western influences.
These distinct aspects of the self emerge from collectivism and individualism thus
“influencing who we are and what we do” (Participant J2). According to participants J2 and
J4 we possess aspects of ourselves that dictate that South Africans are collectively oriented,
due to a focus on family and community, and the progression of a communal united nation.
Conversely, however, “society dictates that we must be an individual” (Participant J4).
Participant J2 further exemplified this fusion in personality to be “because we are embracing
Westernisation”.
Participant F5 contrastingly argued that South Africans possess a more collective orientation
as a result of one’s culture group. This participant (F5) suggested that it is within one’s
cultural group that one’s individual traits emerge; “In South Africa we have more group
personalities than individualistic personalities. I think once you are in your group that is when
you try to find where you belong within that group”. Participant F2 further explored the issue
of collectivism and individualism in personality explaining, “America has become a symbol
of Western culture” but “small town America is not individualistic, that is very community
[based].” This participant adds that in China “people are group orientated, but you will find
that mostly in the rural areas” and as a result of communism. Thus, participant F2 continues,
“there are elements of Western culture in South Africa and there are elements of a very
88
collectivistic culture. It just depends on where you are and also the community that you are
in.”
3.8.3.1. Urban vs Rural
The type of community (rural or urban) in which people live is believed, by focus group
members, to affect one’s personality; as the community in which they live dictates the degree
to which one’s cultural group or individuality define their personality and who they are.
The majority (n=11) of participants indicated the effect that living in an urban and rural area
has on one’s culture and personality as a contentious issue. Participant F2 explained that “A
Zulu person from [KwaZulu Natal] is very different from a Zulu person from Diepkloof,” just
as participant J8 stated that “a Pedi in Polokwane [is] different from being a Pedi in
Jo’burg… [because in] those rural parts those values are still very strong.” Participant F3
clarified that cultural groups possess a collective identity but “the personality traits are
different in terms of where people grow up, even though they may share the same race or
cultural background.” According to participant J8, those who live away from home “adopt
that Jo’burg lifestyle and leave some things that [they] were taught.” Participant J9 portrays
this lifestyle by the clothes one wears, explaining that in her culture females are only allowed
to wear skirts, however when she is in the city (Johannesburg) she wears the clothes that are
an external expression of her personality without fear of reprisal. Participant J8 argues that
“an urban area actually allows you to think, you are not put into a little box.” Thus one is
allowed to be themselves and not succumb to the expectations and rules outlined by one’s
culture.
Participant F5 discussed the difference in individualism and collectivism between different
cultural groups within South Africa. This participant (F5) revealed that no culture is more
individualistic than collectivistic or vice versa, but rather that this “just depends on where you
live…. [and] isn’t based on whether you are white or black.” This participant explains that
there is not much difference between a Black and a White person who, for example, both live
in the inner city. Where one lives and their consequent circumstances dictate the degree to
89
which they are individualistic or collectivistic in nature and not entirely what cultural group
they come from.
Participant J3 explains the effect of urbanisation on a national identity within South Africa;
“Urbanisation brings forth Westernisation and Westernisation doesn’t embrace Ubuntu.” This
is argued to be as a result of the walls one builds between neighbours and the issues of safety
that constantly plague our daily lives. “There are a lot of things that stand in the way of that
‘embraciveness’ and I think that move for Ubuntu within us is still there, but just the nature
of life today doesn’t allow that.”
3.8.4. A National Identity
Both focus groups indicated the existence of a communal and national identity unique to
South Africa. This identity was described by some as Ubuntu, while others were not able to
ascribe a name to capture the essence of personality in South Africa.
The concept of Ubuntu was expressed by the majority of participants (n=14) to be a uniquely
South African concept, meaning “Community. We are the community,” stated participant J6.
Participant J3, however, indicated that “the definition of community varies from culture to
culture and people to people.” This was indicated by participants (n=6) to include one’s
family, the people who live on the same street, those of the same race or everyone in general.
Participant J8 claimed “Ubuntu is different, even in South Africa it is different.” This, as
described above, is a result of the way different cultures define the concept of ‘the
community’. Participant J8 provided the example of funerals to illustrate this point; “For
some cultures I think Ubuntu is if you have a funeral everyone must come, but with White
people I noticed a funeral is small and it is so intimate”.
Participant F4, on the other handed, argues that personality in South Africa emerges from
one’s nationality; “I think it is pride; you are proud of where you come from.” This
participant (F4) reasoned that you “identify with your nationality of your country because
90
that kind of shapes you”. This pride and spirit of Ubuntu was displayed in the soccer World
Cup in South Africa in 2010, relayed participant J1. Participant J3 continued arguing “we
have to stretch the borders of Ubuntu passed that… my brother is your brother and my sister
is your sister. Everyone is one.”
Participant J1 suggested the existence of a South African personality as a result of our
multicultural society we embrace other cultures and are thus friendlier than people in other
countries. “I think [Ubuntu] kind of originated here, because we are all so different and yet
we are still trying to create a similar identity, that of South Africa,” stated Participant J1.
Participant J6, on the other hand, expressed that Ubuntu is not only a South African thing, but
that a sense of community exists in other Mediterranean cultures and that “maybe it was just
an old way of surviving”.
Participants (n=2) of focus group 1 indicated the relevance of one’s generation in their
understanding and embrace of Ubuntu. “I think to the younger generation it means something
very different to the older generation,” stated participant J1. This participant (J1) indicated
that the younger generation “have kind of lost what the spirit of Ubuntu is supposed to be” as
we are preoccupied with a more individualistic living characterised by independence and
autonomy than the older generation. Participant J9, on the other hand, indicated that the older
generation lacked this spirit of community for all races but their own due to the effects of
Apartheid.
Participants (n=3) of focus group 1 expressed the South African personality to be an
amalgamation of other cultures as a result of South Africa’s history of Apartheid. Participant
J2 stated, “we borrow a lot from other cultures and other countries just because of our past”.
Participant J3 continued with this expressing that during Apartheid certain personality traits
in South Africa were associated with “Blacks” or “Afrikaans people” for example. Participant
J5 argued that a lot of who we are today is a result of the media and the West, as we have
begun to adopt the ‘American lifestyle’. However post-Apartheid, according to Participant J2,
“it is changing rapidly and we are forming our own identity and our own little nuances and
91
things that would describe us, which can influence us as individuals.” Participant J3 qualified
this explaining that this integration between South African personalities is the consequence of
a bid for unity post-Apartheid; in which the “barriers that used to be there in the way that we
could interact with each other are starting to be removed and starting to become one.”
Conversely, participant J6 stated “I think we are pretty unique compared to the rest of the
world in the sense that we do have [a lot of] different cultures… and we are able now to adapt
to each of those cultures.” This, according to this participant (J6), is what “forms our ‘South
African’ personality.”
The majority (n=16) of participants acknowledged that there are defining characteristics
within personality in South Africa that set it apart from the personalities of other countries
and cultures. Although participant J4 states, “fundamentally we are basically all the same, but
we are different in a way because of our social background or the way we were brought up.”
What sets South Africa apart from the rest of the world culturally, according to participant
F2, is that “in South Africa we are forced to interact with people who are [different to us].”
Participant F1 explains this difference stating, “Our characterisms are different”- the South
African humour, the way we see and do things is different to other countries. This participant
(F1) qualified this, reasoning that “the way [we are] brought up is different.” The South
African lifestyle, according to participant F1, is one that others cannot relate to; the lifestyle,
morals, beliefs and different cultural groups is what differentiates us from the rest of the
world. Participant F4 extended on this declaring that “there is a certain way [of doing things]
in each country and… your personality can adapt.”
According to participant J4, a country’s power, politics and economic standing affect their
culture and consequent personality- “All those factors change what we are and in a way that
is what makes us [all] unique”. South Africa is unique “because of Apartheid, because we
came out of it, because of Ubuntu and because we now live in that world of community… It
is us, we are new, we are the community, which makes us different to everybody else.”
Participant F2 stated, “We are able to laugh at ourselves… It’s kind of everyday things so it
does distinguish us in how we approach things.” Thus it was indicated that there is not only a
cultural barrier, but also a language barrier between South Africa and the rest of the world,
92
“We used to have a few American friends and they obviously never got a lot of what we were
saying”. It therefore seems, according to these participants, that the way different countries
and cultures speak and express themselves affect their personality. Participant F4 explored
the South African personality revealing “I think you identify with being South African.”
According to this participant of the focus groups, “it doesn’t really matter which [cultural]
group you are from, but because you are both South African it is part of your personality.”
3.8.5. Culture
Different cultures are believed to possess different norms and behaviours that dictate
acceptable behaviour. Thus some participants within the focus groups indicated that South
Africa’s cultural diversity is reflected in personality within South Africa; while some
participants indicated a similarity in personality across cultures, particularly Black cultures.
Participants of the focus groups (n=6) indicated a difference in personality between the
different cultures within South Africa. “Different cultures do different things and yes, our
cultures do influence our personalities,” stated Participant J7. This participant (J7) qualified
this statement by saying, “Like in my culture, Swati, women are not inferior [to men], but it
is how you conduct yourself around other people.” This participant believed that one’s
culture dictates how one is supposed to behave within that culture and so “in a way it
influences your personality”. Contrastingly, participant J5 argued that “the attitude and the
opinion is almost all the same in most Black cultures”. This participant (J5) explained, “like
the Sotho’s and Pedi’s are somewhat very similar in treating women that they are inferior to
men”. Thus the mind set and consequent personality of Black cultures are similar in some
respects. Participants (n=2) of focus group 2 stated that personality is not differentially
influenced by one’s culture. “I don’t think your language or the culture that you are in
necessarily reflects your personality, but I think it reflects your opinion of those specific
groups,” stated participant F6. Participant J1 continued with this thought expressing that
culture influences one’s beliefs and ideas which in turn influence personality.
Participant F5 revealed that within South Africa “what binds us together is that we strive for
the same things. Our morals are basically the same but our approach is what sets us apart.”
93
Therefore different cultural groups are similar in personality with regard to the fundamentals;
however there are small differences that set them apart (Participant F5). Participant F6, on the
other hand, rationalised that one’s subgroup dictate similarities in personalities as “the
subgroups are the same and bring similar personalities together”. However, “When you look
at the bigger picture, like in all your groups, that is where the personalities differ completely”
(Participant F6).
Participant F2 reasoned that “no one has the same personality” but “You do find certain traits
and characteristics are more common [in] certain groups”. Participant F2 clarified this saying,
“You tend to hang around with certain people that are similar to you.” Thus if you are Zulu
and are surround by Zulu family and friends these ‘Zulu’ traits emerge, influencing and
moulding who you are. This argument thus provides evidence for the Five Factor theory and
trait approach to personality, indicating overarching traits that describe all individual
personality.
Participant F5 explored the reasons that one may view people of the same culture to possess
similar personality traits. It was argued that in order to understand people we create
“stereotypes and generalisations” based on our interactions with people from that cultural
group. This allows us to “come to some sort of conclusion” about them in an attempt to get to
know them; however these stereotypes may be adapted as there is “space for change”.
Participant F3 agreed with this stating that we place people into classifications and identify
them with certain ideas or traits as “It is easier for people to make understandings about those
different cultures”. This participant (F3) reveals a major role player in this process of
classification saying “I think the media also plays a very big part in how people understand
certain cultures and how they are portrayed on TV.”
3.8.6. Psychometric Testing
94
Participants identified a number of common issues in the SAPI and CPAI-2 with regards to
psychometric testing.
3.8.6.1. Length
The first issue explored the length of the SAPI and CPAI-2 questionnaires. The amount of
time taken to fill out a psychometric test reflects the quality of the answers presented by
participants. The longer the test, the less attention is paid to the items and their answers, as
participant J6 reveals “we just wanted to get it over and done with and so we didn’t put much
thought into it”; thus this leads to “the fact that we wouldn’t answer it so truthfully”.
Participant F9 elaborates saying, “People don’t answer properly- they don’t care. I saw some
people just ticking to get it over with.” In contrast, however, participant J4 argues that this is
dependent on the person and how interested they are in completing the questionnaire. “When
I was doing it, yes it was long, but I made sure that every question was truthfully answered to
what I wanted as me” (Participant J4).
Participant J4 suggested a solution to the length of the questionnaires and possible resulting
consequences; revealing that shortening the length of the scale will minimise the length of
time required to complete the questionnaire. “I think because the questionnaire was so long
you got bored with it”, therefore although a yes or no scale limits the accuracy and depth of
one’s rating of an item it takes less time to fill in the questionnaire so “at the same time it can
work for and against you” (Participant J4).
3.8.6.2. Repetition
The second issue explored the repetition of some of the items of the SAPI and CPAI-2
questionnaires. The length of the questionnaire is considered by participants (n=5) to be a
result of the repetition in items. This impacted on the experience of the participants when
filling out the questionnaire, as participant F8 revealed “I was very bored. I can’t remember
what one of the questions was but it was asked in one way here and asked in a different
waythere.” Participant J4 believed this to be a good thing, as if one provides the same answer
then “it is a true reflection” of their personality. Participant J2 revealed that although
95
questions were repeated and worded differently in the questionnaire and could have elicited
different responses, “I still think they would be in a similar range in terms of honesty.”
Participant F8 qualified this by saying that although some of the questions were “very broad
and didn’t specify what was being asked… my understanding of both of them were exactly
the same.” Participant F6 revealed that rewording the items made you really think of the
answers that “you take for granted day to day”. Participant F5 took a different stance
revealing that the reworded questions at times did result in different answers. “I did answer
differently because some of the things like ‘I love people if I don’t get along with them’,
sometimes I was like ‘I don’t love them, so no’ and then sometimes I was like ‘well I can be
mature about something so yes’”. Participant J3 elaborated on this revealing that the answers
to the questionnaire depended on how the person is feeling on the day that the questionnaire
is completed.
Focus group 2 raised the issue of mood when completing a long and repetitive questionnaire.
“I think your mood affects you very much,” stated participant F1. Participant F10 agreed
saying, “Ja, by the end of it you just want to finish it.” This participant (F10) continued
revealing that in the beginning of the questionnaire participants are aware that it is long,
therefore they rush it. However, later on when they see certain questions are repeated it
makes them stop and think about how they answered it the first time “and how it changes
your thought process between the two.”
3.8.6.3. Rating Scale
The third issue explored the use of rating scales. The SAPI and CPAI-2 adopt a 5-point rating
scale of their items. Participant J4 stated that the rating one would give to an item “is
dependent on what context you were actually looking at the question”. According to J4, the
context in which the item is presented will affect how the participant responds. Thus the
participants rating may be different on another day due to their interpretation of the item at
that time. This, according to participant J4, could result in a neutral score to be safe and avoid
misinterpretation. Participant J4 also revealed that ratings may change because at “certain
times maybe I won’t be so open” or I may answer differently “when I am around other
people.” Participant J3 suggested in order “to minimise that margin of error you could just
96
reduce the scale from one to five maybe to just one to two. It will make it more direct because
I will not pick the middle one. It is a quick option out.” Participants (n=3) acknowledged that
this would be forced choice and thus argued for a larger rating scale.
Participant J4 revealed that when answering questions with conclusive terms such as
“always” or “never”, the scale options of ‘agree’ and ‘strongly disagree’ helped to answer
more accurately. Participant F1 suggested using a more qualitative open-ended method would
allow for “a more diverse description of people and they can actually think about themselves
and answer truthfully by writing it down.” Participant F4 believed that “there should be an
option of ‘it depends’” in order to account for the ambiguous and confusing items.
3.8.6.4. Problematic Items
3.8.6.4.1. Linguistically Problematic Items
The fourth issue explored linguistically problematic items of the SAPI and CPAI-2.
Participants of the focus groups found a number of problems in the way that the items were
phrased. Participant J4 revealed that the use of the word “some”, as in the item “I have some
bad habits” is a bit vague and ambiguous. It is “not specific, like what are ‘bad habits’? You
might see smoking as a bad habit, but I don’t care”. Thus, according to participant J4, the
answer to such an item will depend on the interpretation and how you view your reality and
therefore will not be consistent amongst all participants. Participant, J1, on the other hand felt
that specific terms such as “usually”, “always” or “never” were also problematic. “You have
to think to yourself ‘well I did that once, but does that still count as never?’” Participant J1
revealed that it was, therefore, difficult to strongly agree or disagree with items using such
conclusive terms.
Participants (n=5) identified a number of linguistically problematic words used in the items
of the questionnaire; namely “etiquette”, “thrifty” and “stigma”. Participant J6 felt that a
large part of the population would not understand words such as these. Participant F2
indicated that “filial piety” would be difficult for the average person to understand and also
proved to be problematic for many participants when actually filling out the questionnaire.
97
Participant J3 identified context in the use of words and phrases to be problematic, such as
“cohabitation”. Participant J3 elaborated by saying “If you ask me if I think I am better than
other people, if I am being bold I will say that I think I am better than them in running for
sure! So in what context?”
Participant J2 found the way that some of the items were phrased or worded to be
problematic, “I don’t think the language itself in terms of the words [is problematic]”.
Participant J1 argued that one’s up bringing would be of importance “and the way you speak
at home”. For example “I would speak more like ‘I guide people to make the right decisions’
instead of ‘I teach people to make the right decisions’” (Participant J1). This participant (J1)
felt that this could therefore contribute to the way one interprets and responds to certain
items. Participant F8 declared the item “I hug people when I am happy about them” to be
problematic as it is ambiguous; “Like you are happy about what? Seeing them?”
Focus group 2 raised the issue of relevance in respect to some items in the questionnaire.
Participant F1 declared that “Some of the questions were odd.” This participant (F1) felt that
items such as “I frequently buy snacks” are confusing and seemingly irrelevant to the topic of
personality under investigation in the questionnaire- “like what does food have to do with
personality?” Another problematic item according to participant F1 is “’if someone cooks
supper for me I will return the favour’, but I suppose that is like if someone does something
for you then you will return the favour.” It seems that the wording and intention of certain
items are not always congruent for the participants.
Participants (n=2) in focus group 2 revealed that the use of English expressions could also be
problematic for those individuals who do not have English as their first language. Participant
F4 explained that the expression “blow their trumpets” is difficult to interpret. Participant F4
believed that this could mean a boost of confidence, while participant F2 interpreted this to
mean giving someone praise.
Participant F1 suggested that such items should be simplified so that the meaning is not
ambiguous or confusing but consistent for all. Participant F6 continued on this to suggest that
98
colloquial language should not be used as “my tone and style of speaking could be very
different from everyone else’s’.” This participant (F6) felt that all participants of all cultures
and personalities should be able to understand the items.
Participants (n=3) of the focus group 1 noted that level of education of the test-taker will
influence their ability to successfully complete the SAPI and CPAI-2. One’s level of
education will dictate their capacity to understand the items presented in this questionnaire.
Participant J2 explained that the items may be culturally or linguistically problematic “if I
had to sit my gardener down”. Participant J2 continued “if you have got intellect you can
rationalize and figure anything out… for a big part of our country this would be a problem
because I don’t think they would understand the concepts”. Participant J1 qualified this by
saying, “It could have to do with the school you went to”.
3.8.6.4.2. Culturally Problematic Items
The fifth issue explored culturally problematic items of the SAPI and CPAI-2. Certain items
were found to be culturally specific by participants (n=9) in the focus groups in that they
made reference to certain cultures so that individuals excluded from the said culture could not
relate to the item. Participant J1 identified an item in relation to an ancestral sacrifice, “For
me obviously I don’t really know what ancestral sacrifice is, so to think in that context kind
of threw me off.” Participant F6 continued stating that some participants may not be able to
relate to weddings or funerals with ancestral sacrifices. The item “If the man does not earn
the highest salary in the house, he should still be the head of the house”, was found to be
culturally biased by participant J3. “In a cultural context, it is not about who is bringing in the
money, it is about that the money is coming in.”
Participant F3 found the item “even though I am not very rich, I would still prefer to buy the
most expensive out of several brands of the same product” to be culturally significant. This
participant (F3) reflected, “It’s interesting how people view branding in terms of different
cultures.” To this participant (F3) purchasing brands is a means of self-acceptance or to imply
status which varies across cultures; “The understanding of self-acceptance is through other
99
people thinking they have got money or they are cool.” In opposition, participant F7 stated
that this is a result of one’s socio-economic circumstances and personality, rather than
culture- “within a culture you can get poverty and extreme wealth.” Participant F1 declared
that people in Western cultures that “live in the cities obviously have a different perspective
of life and [therefore] they would want to live at a certain standard”. Participant F3 qualified
this by saying, “you will drive in townships and there will be a 5 series BMW… I am not
saying it is only directed at Black people or White people, it just happens quite a lot in terms
of how people brand themselves in terms of clothing and image.” This indicates that the way
that one interprets this item may be a result of where one lives, as well as culture. Participant
F2 describes this item to be of Chinese origin, because “As western culture spreads further
and further materialism becomes a measure of success”, whereas in traditional rural cultures
this is not particularly a priority and success is not measured in such forms. Participant F5
reasoned that traditional rural cultures display such wealth and success in the amount of cows
one owns; however, this participant (F5) acknowledged that even rural cultures are
progressing in how they show their wealth and so it is becoming less and less a result of
culture.
Participant F2 revealed that some of the ideas and concepts within the questionnaire
harboured “a bit of Chinese influence”. An example of this according to participant F2 and
F10 was the item “to maintain a pure and simple culture eccentric clothes and hairstyles
should be strictly banned”. This is believed by participant (n=2) to be a Chinese concept, as it
does not fit with Western tradition or cultures. Participant F3 found the item “One can avoid
serious mistakes by always avoiding tradition” to be culturally problematic as it depends on
one’s cultural traditions and whether or not that view themselves as individualistic or
collectivistic.
3.8.7. Social Desirability
100
A final theme to emerge was that of social desirability. Participants (n=3) stated that some
items were difficult to answer completely honestly as the items were sensitive in nature or
had possible negative connotations; for example, “Do you view yourself as better than other
people?” as indicated by participant J3. Similarly the item “Do you accept people’s
apologies?” was also found to be problematic, as when answering these items “you don’t
want to sound too vain or too full of yourself”, nor do you want to feel like a horrible person
(Participant J3). People will answer in such a way as to portray themselves in a good light
“and say ‘No, I am not rude. I am direct!’” (Participant J3). Participant J1 stated “it is kind of
an ego thing, where we kind of get taught what we shouldn’t say in public”. However,
participant J2 argued that social desirability is affected by one’s circumstances and
experiences; “If I had done this ten years back I would have done it totally differently. I can’t
see myself as better than someone because that would be seen as arrogant.” According to
participant J2, now “what I have experienced… has drastically altered my perception of
things”.
Participant F4 revealed that items such as “I am a very nice person” are difficult to answer;
“it’s like ‘well am I? ...how do you answer that yourself?” Participant F7 explains, “I don’t
think you get the opportunity to judge yourself to that extent.” Thus, according to participant
F4, items such as “my friends say I am [a nice person]” are easier to answer.
Participant J2 stated “when I did it [the questionnaire] certain questions were a direct
stimulus for me… into more specifically pain oriented thoughts in answering the question”.
This participant (J2) revealed that the item, “There is no stigma attached to marrying a
divorced person” was especially personal, as according to this participant there is a stigma
attached to divorce and it is almost seen as wrong. “[I]f I had answered it ten years back
when I wasn’t married, then I probably would have gone ‘ag, so what’” (Participant J2).
Participant F7 felt that this item was badly worded and would thus be difficult to answer.
“[T]he question doesn’t state ‘what do you think?’, like we are saying is there a general
stigma about it?”
Participant F6 revealed how the wording of repeated items may result in participants
answering in more socially desirable ways. According to this participant (F6), when two
101
questions are worded the same, “one can be worded with more positive connotations so you
tend to answer it differently.” Participant F6 continued, “Like if it says ‘You make people
feel uncomfortable’ and ‘You make people feel stupid’; ‘Uncomfortable’ is more of a
positive connotation type of thing in comparison to the other one. So I think you let your best
part shine through”. Participant F5 indicated that “the fact that we started answering these
questions in class made more people answer it. I mean you don’t want to be the baddie that
doesn’t answer the questions, meanwhile everybody is doing it.” Consequently, according to
this participant (F5), participants just randomly ticked to avoid being “the loser that walks
out”; thus indicating a social pressure to complete the questionnaire.
Participant F3 revealed that items such as, “I insult people” realistically require a “yes”
answer as “even though I do them… I don’t do them all the time”. However “I don’t want to
be one of those people that give answers that aren’t good” (Participant F3). Participant F3
revealed that the way one answers the questionnaire depends on “honesty and the degrees at
which you know yourself”.
3.9. Conclusion
This chapter has provided all the results relevant to the study; beginning with descriptive
statistics for the sample, followed by descriptive statistics for the two instruments- the SAPI
and CPAI-2. After this, the results pertaining to the reliability, construct validity and
construct bias of these two instruments.
The descriptive statistics for the sample were considered to be adequate for gender, race and
home language; after collapsing the groups for race and home language into Whites and nonWhites, and English first language and English second language speakers respectively. The
descriptive statistics for the SAPI and the CPAI-2 were also found to be adequate, with all the
scales and subscales being normally distributed.
Results pertaining to the internal consistency reliability of the SAPI and CPAI-2 were then
reported. Internal consistency reliability was found to be appropriate for all the clusters and
102
sub-clusters of the SAPI, except for the Social Desirability cluster and sub-clusters.
Furthermore, internal consistency reliability was found to be appropriate for three of the
CPAI-2 scales- Traditionalism vs Modernity, Social Sensitivity and Harmony, as well as the
Interpersonal Relatedness factor and problematic for three CPAI-2 scales- Ren Qing,
Discipline and Thrift vs Extravagance.
Construct validity results using factor analysis with varimax rotation revealed that the SAPI
did not demonstrate appropriate construct validity, with none of the clusters and sub-clusters
replicating as they should. However, the three factor solution at scale level appears to provide
the best fit, with Social Desirability and its sub-clusters loading appropriately. Furthermore,
the item analysis revealed that the items of the SAPI do not load as expected for the most
part. It appears, in addition, that items from the Soft-Heartedness and Relationship Harmony
sub-clusters combine to form two factors. Construct validity results using factor analysis with
varimax rotation for the CPAI-2 demonstrated appropriate construct validity, with the six
scales of the Interpersonal Relatedness factor replicating as proposed by Cheung et al. (2008).
The two factor solution, on the other hand, was found to problematic, particularly for the
Thrift vs Extravagance scale. A joint factor analysis of the SAPI and CPAI-2 collective
dimensions did not replicate according to the proposed structure and therefore did not
demonstrate adequate construct validity.
Construct bias results in terms of gender, race and home language were then presented,
generally indicating that Egoism, Hostility, Active Support, Empathy, Facilitating, Guidance,
Encouraging Others, Integrity, Approachability, Harmony Maintenance, Harmony Breach
and Impression Management of the SAPI demonstrate evidence of construct validity; while
Interpersonal Relatedness, Ren Qing, Discipline, Social Sensitivity, Harmony and
Traditionalism vs Modernity of the CPAI-2 demonstrate evidence of construct validity.
This chapter concluded with thequalitative thematic content analysis results of the focus
groups. The themes that emerged included a general understanding of personality, the innate
existence of personality vs the over expression of personality, collectivism in personality, a
103
national identity, culture, psychometric testing and social desirability.These results will be
discussed further in Chapter 4.
104
Chapter 4: Discussion
4.1. Introduction
This chapter offers a discussion of all the results obtained in this study in accordance with the
aims of this study. Therefore the descriptive statistics for the sample of this study are
discussed first, followed by a discussion of the descriptive statistics for the SAPI and CPAI-2.
The discussion will then follow on to include the results pertaining to the internal consistency
reliability of the two instruments under investigation, and subsequently a discussion of the
results pertaining to construct validity. Following from this, all results pertaining to issues of
construct bias are discussed. The results of the qualitative thematic content analysis will be
discussed. The chapter concludes by addressing the implications of these results for
psychometric research in South Africa, the etic-emic debate and debates on national identity.
4.2. Descriptive Statistic: Demographic Information
The total sample size of this study was 489 first-year Psychology students. This sample size
can be considered fairly large and is therefore deemed to be adequate when considering the
aims of the study and the statistical analyses conducted on the entire sample. However, when
examining construct bias in terms of gender, race and first language; the sample was
examined in terms of sub-groupings. From the demographics statistics sample Table 3.1 it is
evident that gender, race and first language groupings were unequal, with a predominance of
a particular group in each category i.e. female, White and English respectively. The sizes of
some of the subgroupings for race and first language were considered too small to conduct
statistical analyses; hence the subgroupings for these demographic variable where collapsed
into White and non-White, as well as English first language and English second language, as
stated in Chapter 3.
From Table 3.1 it was evident that even English second language speakers (n = 169) largely
rated their English reading ability as ‘good’ or ‘very good’, with only five participants rating
their reading ability as ‘very poor’ or ‘poor’. Thus differences found in terms of the variable
105
first language can be considered to be actual true differences not differences due only to
language proficiency.
The gender distribution of the sample in this study indicates a large female majority (n = 373)
of 78.20%; similarly the University of the Witwatersrand student demographics reveal that
52.76% of the student population is female (Wits University Strategic Planning Division:
Facts and Figures, 2010). However, it has also been suggested that psychology students of
this University tend to be predominantly female (Laher, 2010). As the questionnaire was
completed during lecture times, issues with regard to attendance may have affected the
representativeness of the sample. Similarly those individuals that remained in the lecture to
complete the entire length of the questionnaire will likewise affect the attendance and
consequently the representativeness of the sample. The distribution of age was positively
skewed, with the sample ranging from 17 to 44 years of age, with a mean of 19.1 (SD= 2.08).
Once again this is congruent with what would be expected to be the age of first-year students
at the University of the Witwatersrand.
According to the most recent mid-year population estimates for South Africa (Statistics South
Africa, 2011), approximately 52% of the population is female. The population group
estimates indicate that 79.5% of the population constitutes Africans, 9% of the population
Coloured, 2.5% Indian/Asian and 9% White; therefore the total non-White population
constitutes 91%, however 55.46% of this population were only represented in this sample.
These statistics reveal the under-representation of non-White population groups in this
sample. Thus the racial distribution of the sample does not adequately represent the racial
distribution of the population and could thus be said to impact the aims of the study.
This under-representation of the non-White demographic in the sample is a result of sampling
difficulties and is also reflective of the persistent inequalities rife in South Africa due to
political dispensation of Apartheid. The limited educational and economic opportunities
afforded to the non-White demographic are a direct result of restricted access to higher
education (Sehlapelo& Terre Blanche, 1996; Laher, 2010). Although universities within
South Africa are attempting to redress these inequalities, the sample in this study reflects that
106
education is still considered a privilege for the non-white population despite the
improvements in racial representation post-apartheid (Laher, 2010).Consequently, the size
and composition of the sample limits the generalisability of the results obtained in this study.
However, it is imperative to emphasise the importance of the contribution of this study to
transformation within South Africa revealing data with regards to equality and opportunity.
The value in this study thus lies in its exploration of personality psychology within South
Africa, therefore contributing to local and international research.
4.3. Descriptive Statistics: SAPI
Descriptive statistics in the form of means, standard deviations, minimum and maximum
values, and skewness coefficients for the clusters and sub-clusters of the SAPI are presented
in Table 3.2. This Table suggests that all the clusters of the SAPI were normally distributed.
The means, standard deviations, and maximum and minimum values obtained in this study
appear to generally be within an acceptable range. However no descriptive statistics were
reported by Nel (2008) from the normative data. Similarly, no other publications report these
statistics as analysis of the SAPI is still underway.
4.4. Descriptive Statistics: CPAI-2
Descriptive statistics in the form of means, standard deviations, minimum and maximum
values, and skewness coefficients for the Interpersonal Relatedness scale and subscales of the
CPAI-2 are presented in Table 3.3. From this Table it is evident that all the CPAI-2
Interpersonal Relatedness factor and scales are normally distributed.
An examination of the results in this study as compared to those obtained by Cheung et al.
(2008) and Laher (2010) are not possible due to the different rating scales utilised for each
study. Cheung et al. (2008) and Laher (2010) utilised a true-false format for the CPAI-2,
while in the current study a 5-point rating scale was utilised in order to achieve uniformity
with the SAPI (V.H. Valchev, personal communication, December 14, 2011). This is a
drawback of this study as such a comparison would have provided valuable information on
107
the collective dimension Interpersonal Relatedness in a South African sample. It also would
have allowed for a comparison to the normative Chinese sample- this would have allowed an
examination of the universality of the collective dimension of personality.
4.5. Internal Consistency Reliability: SAPI
The internal consistency reliabilities for the SAPI clusters and sub-clusters all range from .62
to .96; with the exception of the Social Desirability clusters and sub-clusters. The Social
Desirability cluster had a coefficient of .53, while the Social Desirability sub-clusters range
from .17 to .33 (see Table 4.1).
The Social Desirability cluster in the SAPI is made up largely of items from the MarloweCrowne Social Desirability Scale (MCSDS) and the Balanced Inventory of Desirable
Responding (BIDR). The Marlowe-Crowne Social Desirability Scale (MCSDS) intends to
measure a single construct- need for approval- through infrequent but socially approved
behaviours, as well as frequent but socially disapproved behaviours (Uziel, 2010;
Leite&Beretvas, 2005). The BIDR, on the other hand, measures two factors: self- deception
and impression management. Self-deception is defined as the tendency to give honest but
positively biased responses; while impression management is defined as the habitual
presentation of a favourable public image respectively (Paulhus, 2002; Uziel, 2010;
Leite&Beretvas, 2005). Impression management assesses one’s overt behaviours whose true
nature are arguably accessible to the individual, such that they can either lie or be honest
about them. Self-deception, on the other hand, assesses potentially psychologically
threatening thoughts and feelings that appear arguable less accessible to the conscious mind,
resulting in possible denial or unconscious defenses (Uziel, 2010).
In light of these descriptions and definitions, it appears at a surface level that the Social
Desirability scale of the MCSDS and the Impression Management scale of the BIDR may in
fact be measuring similar constructs, as one’s need for approval and desire to obtain a
positive image are related. Further research into this required. Furthermore, one is compelled
to notice the surface level connection or overlap between the construct of social desirability
108
and that of Integrity proposed by Nel (2008). Integrity, as described by Nel (2008, p. 133), is
“the moral soundness of a person or the tendency to be trustworthy and to act in an honest,
pure, and responsible manner.” The clear connection between these two constructs thus
provides insight into the overlap found between the Social Desirability sub-cluster and
Integrity of the SAPI in subsequent analyses, discussed later in this chapter. It appears
therefore that a problem particularly exists with the MCSDS component of the Social
Desirability scale; however low reliability coefficients were found for all the sub-clusters of
this cluster.
The difficulty with social desirability emerges in the way that we define and particularly label
the components of this concept, especially when using them interchangeably (Uziel, 2010).
The scale construction and composition of the Social Desirability cluster of the SAPI may
therefore be the reason for the less than satisfactory reliability coefficients of this cluster and
its sub-clusters. Theoretically, the Social Desirability cluster of the SAPI comprises of the
three sub-clusters Social Desirability, Impression Management and Self-deception. The
distinction between all three sub-clusters therefore becomes questionable as Impression
Management and Self-deception have been theorised by many researchers and practitioners
over the years to be components of Social Desirability (Uziel, 2010), consequently causing
one to question what the Social Desirability sub-cluster is assessing that is different to the
combination of Impression Management and Self-deception sub-clusters. The lack of clear
dimensionality for social desirability scales has been a concern since its development (Uziel,
2010) and appears to remain a concern for this instrument. Another reason that could be
provided for the low reliability coefficients found for the Social Desirability cluster is the
lack of statistical evidence to support the view of social desirability scales as a reliable
validity measure or one that detects bias reporting (Piedmont, McCrae, Riemann
&Angleitner, 2000; Uziel, 2010). Research on the MCSDS in particular, has indicated that
Social Desirability as measure in this scale is regarded as a personality trait more than a
response bias (Verardi et al., 2010); thus revealing the potentially problematic use of this
scale as a social desirability bias.
On examination of the items that comprise the Social Desirability sub-cluster of this
instrument, it appears as if they should in fact form part of the Impression Management or
109
Self-deception sub-clusters. Items such as 28- ‘I think about my options before I make a
choice’- and 84- ‘I have doubts about my ability to succeed in life’- can be considered in
terms of the definitions of social desirability provided above, such that an argument could be
made for such items to form part of the Impression Management and Self-deception subclusters respectively. This may explain the higher reliability (.33) found for the Social
Desirability sub-cluster as compared to those found for the other two sub-clusters. Further
research into the items that comprise these sub-clusters is therefore warranted, such that poor
items are revised.
A frequent criticism of social desirability dictates that it “is a source of bias only to the extent
that it represents respondents’ conscious deception efforts” (Uziel, 2010, p. 244); thus placing
doubt on the self-deception scale as an accurate measure. Consequently this criticism may
provide an explanation as to the reasons for the Self-deception sub-cluster reliability of .20;
as this figure would presume that this sub-cluster is unreliable as a measurement. Further
work on the Social Desirability cluster of the SAPI is required, coupled by further research
and validity and bias evidence in order to determine the source of the problem.
Table 4.1 presents the reliability coefficients for this study as well as the normative sample
for the SAPI clusters. The normative data for the SAPI sub-clusters are, however, not
available as the developers of this instrument are still working on these sub-clusters.
The Soft-Heartedness cluster in this study achieved a reliability coefficient (.86) identical to
that of the 59 item normative sample for this cluster. Similarly, the Relationship Harmony
cluster in this study revealed a Cronbach Alpha of .81, identical to the 48 item normative
sample for this cluster, as indicated in Table 4.1. Although the reliability coefficients for
these clusters are comparable to the normative sample, it is important to note that this study
utilised more items for the Soft-Heartedness and Relationship Harmony clusters, with 100
and 76 items used respectively. Consequently the number of items utilised should have
allowed for higher reliability coefficients in this study, because as Carmines and Zeller
(1979) suggest- the number of items increases so the value of alpha also increases; however
“adding of items indefinitely makes progressively less impact on the reliability” (Carmines
110
&Zeller, 1979, p. 46). This thus indicates that not as many items are needed for the subclusters. Therefore further research is required; particularly into whether the inclusion of
additional items adds anything substantial in terms of reliability.
Table 4.1: Cronbach Alpha coefficients for the SAPI scales
SAPI Scales
Soft-Heartedness
Egoism
Hostility
Empathy
Active Support
I help others to
succeed
Facilitating
Guidance
Encouraging Others
Integrity
Integrity
Relationship
Harmony
Approachability
Harmony
Maintenance
Harmony Breach
Social Desirability
Social Desirability
Impression
Management
Self-deception
This study- α
0.86 (100 items)
0.62
0.90
0.92
0.90
0.96 (55 items)
0.96
0.86
0.63 (34 items)
0.63
0.81 (76 items)
Velichko Pilot
Study- α
0.86 (59 items)
0.89 (78 items)
N/A
0.97 (55 items)
0.61 (34 items)
0.80 (51 items)
0.81 (48 items)
0.71
0.93
0.82
0.53 (25 items)
0.33
0.17
0.20
The reliability coefficients for the Facilitating and Integrity clusters in this study were found
to be almost identical to that of the normative sample, presented in Table 4.1. The reliability
coefficient (.96) for the Facilitating cluster in this study (55 items) is slightly smaller in
comparison to that of the normative study (.97); while the coefficient for the Integrity cluster
(34 items) in the current study was found to be slightly higher than the normative study, with
internal consistency reliabilities of .63 and .61 respectively.
111
A reliability coefficient was not available for the ‘I help others to succeed’ sub-cluster, as this
sub-cluster is comprised of one item. Single item scales produce responses that are not
consistent over time, therefore rendering them unreliable. It can also be said that the
characteristic measured in this sub-cluster is broad in scope, thus requiring several items to
assess it (Spector, 1992). This consequently indicates that this sub-cluster needs to be revised.
The SAPI clusters and sub-clusters are still under construction, thus there are no internal
consistency reliabilities to compare for the SAPI collective dimension sub-clusters. The lack
of available data for the reliability coefficients of the SAPI sub-clusters in the normative
sample is problematic. Consequently, these figures need to be obtained as well as further
research conducted in order to investigate the internal consistency reliability of this
instrument in the South African context, as well as to obtain a better understanding of the
adequacy of the scales and subscales of the SAPI.
4.6. Internal Consistency Reliability: CPAI-2
Table 4.2: Cronbach Alpha coefficients for the CAPI-2 scales
CPAI-2
Interpersonal
Relatedness
Traditionalism vs
Modernity
Ren Qing
Social Sensitivity
Discipline
Harmony
Thrift vs
Extravagance
This study- α
0.79
Cheung et al. (2008)
0.70 (+0.09)
Laher (2010) α
0.74 (+0.05)
0.66
0.65 (+0.01)
0.69 (-0.03)
0.44
0.60
0.57
0.68
0.36
0.52 (-0.06)
0.62 (-0.02)
0.59 (-0.02)
0.51 (+0.17)
0.57 (-0.21)
0.39 (+0.05)
0.45 (+0.15)
0.60 (-0.03)
0.56 (+0.12)
0.37 (-0.01)
The internal consistency reliability for the CPAI-2 Interpersonal Relationship factor in the
current study, the normative study by Cheung et al., (2008) and Laher’s (2010) study are
presented in Table 4.2. The differences in reliability coefficients for the CPAI-2 factor and
112
scales between the current study and both the normative study and Laher’s (2010) are
presented in brackets in the Table in the respective columns.
Three scales of the Interpersonal Relatedness factor obtained internal consistency reliabilities
higher than .60; while the other three scales were found to be lower than .60 (Nunally&
Bernstein, 1994). The overall Interpersonal Relatedness factor, however, had a reliability of
.79. This was followed by Laher (2010) (.74) and lastly the normative study (.70). This is an
interesting finding in light of the fact that the CPAI-2- and particularly the Interpersonal
Relatedness factor- was designed to assess Chinese Personality from an indigenous
perspective and warrants further examination in terms of validity and bias which will be
presented in the discussion to follow (Cheung et al., 1996; Cheung et al., 2001).
All the reliability coefficients for the Interpersonal Relatedness factor and its subsequent
scales are comparable to the normative sample, as well as Laher’s (2010) study for the most
part. For the Traditionalism vs Modernity scale reliability coefficients range from .65 to .69
with minor differences in comparison to the current study. It is interesting to note, however,
that the internal consistency reliabilities for this scale were higher in the current study, as well
as Laher’s (2010) study (who obtained the highest reliability) conducted in a South African
sample as opposed to the normative Chinese sample.
The highest reliability for the Ren Qing scale was found for the Chinese normative sample,
followed by the current study and then Laher (2010). However poor reliability coefficients
were found across the board for this scale. The normative sample correspondingly found that
Ren Qing did not achieve factor loadings of above .40 (Cheung et al., 2008). The reliability
coefficient for the Social Sensitivity scale of this study just reaches .60; while the normative
study experienced a reliability of .62. In contrast, a large difference occurs between the
current study and Laher’s (2010) study for this scale with a difference of .15.
The Discipline scale reveals comparable reliability coefficients for the current studies, as
compared to the normative sample and Laher’s (2010) study; however only
113
Laher’scoefficient reaches .60, with the normative and current studies obtaining poor
reliability coefficients. Interestingly, the reliability coefficients for the Harmony scale provide
the greatest difference for the current study; with differences of .17 and .12 for the normative
study and Laher’s (2010) study respectively. Thus the reliability coefficient for the current
study is the only one to exceed .60. The Thrift vs Extravagance scale presents the most
problematic reliability coefficients across all studies, revealing coefficients below .60. The
current study and Laher’s (2010) study present comparable internal consistency reliabilities
with a difference of .01, the current study being the lower of the two. When compared to the
normative sample, a difference of .21 exists.
Given the results it can be argued that difficulties may exist in terms of the items that
constitute these scales or their scale construction. Furthermore, these results may reflect
problems with regard to the constructs assessed in the CPAI-2. Consequently these potential
issues need to be further explored in terms of validity and bias. The CPAI-2 collective
dimension appeared to have differential utility in the South African context- with three of the
six Interpersonal Relatedness subscales revealing poor reliability coefficients, while the
Interpersonal Relatedness scale itself had a high reliability of .79. Thus the utility of the
collective dimension in this instrument is not clear, indicating the necessity for further
examination of the subscales and factor structure of this etic measure of personality.
4.7. Construct Validity: SAPI
4.7.1. Scale Level
Construct validity of the SAPI was conducted at scale level using exploratory factor analysis
with varimax rotation. At a theoretical level the SAPI inventory postulates four broad
collective dimension clusters- namely Soft-Heartedness, Facilitating, Integrity and
Relationship Harmony. The SAPI also includes a fifth cluster- Social Desirability, which is
theoretically proposed to be independent of the other four collective dimension clusters of the
SAPI.
114
Consequently a five factor solution was explored including the Social Desirability dimension.
The resulting scree plot and eigenvalues indicated the retention of three factors respectively,
thus a three factor solution was also explored. The five factor solution indicated an overlap
with the Integrity cluster and Social Desirability sub-cluster; while the three factor solution
revealed that the Social Desirability cluster loaded as theoretically proposed. Therefore a four
factor solution was conducted without the inclusion of the Social Desirability cluster. This
four factor solution and its resulting scree plot and eigenvalues indicated the retention of two
factors; consequently a two factor solution was also explored. In the sections to follow, the
SAPI factor solutions including the Social Desirability cluster will be discussed, followed by
the factor solutions excluding the Social Desirability cluster.
It is evident from these solutions that a problem exists in the scale construction of the
collective dimension of personality in this instrument, due to the unexpected loading of subclusters on factors that do not represent overall clusters as they were originally
conceptualised. This problem, however, could be a result of the number of items used in this
inventory, as 265 items were used to assess these four clusters. This may result in redundancy
or repetition in the meaning of items used to measure the same construct. Similarly, the way
these items were constructed and put together may affect the problematic scale construction
of this inventory. The scales of the SAPI were constructed from everyday conceptualisations
of personality across language groups. Therefore it may be said that the South African
personality may emerge or be expressed differently to what was expected, in that different
aspects of personality combine in unexpected ways (Meiring, 2006; Nel, 2008).
The total variance explained by the three, four and five factor solutions was 69.75%, 84.84%
and 81.04% respectively. These figures indicate that the three factor solution may be best to
describe the collective dimension of personality in South Africa assessed by the SAPI. This
also indicates that the four factor solution may be adding more to the three factor solution
than the five factor solution; thus providing some insight into the value of the Social
Desirability cluster. Additionally, the order of the factors in each solution may provide some
important and interesting information as to the collective dimension of personality in South
Africa; yet further research is warranted in order to determine if this order is in fact
problematic. When considering the variance explained by each factor for all the scale level
115
factor solutions, it becomes apparent that Factor 1 and 2 explain the most variance for each
solution; with Factor 1 and 2 explaining 40.29% and 20.40% respectively for the factor
solutions including Social Desirability (i.e. three and five factor solutions), and 50.66% and
21.57% respectively for the factor solutions excluding Social Desirability (i.e. two and factor
solutions). These figures indicate that Factor 1 may be the most important determinants of the
collective dimension of personality in South Africa, followed by Factor 2. This is particularly
interesting as Factor 1 was found to be largely representative of a collectivistic dimension of
personality in South Africa, akin to elements of the concept of Ubuntu; while Factor 2 was
found to represent those individualistic elements of personality that inhibit the collective
dimension and the spirit of Ubuntu. This will be discussed further in the sections to follow.
This therefore provides evidence of the collective dimension of personality in South Africa;
however more work is required for this instrument to be valid in the South African context.
Cross-loadings occurred for the scale level solutions (expect for the two factor solution) pf
the SAPI, however, this is expected as personality dimensions are usually related and
consequently overlap.
4.7.1.1. Factor Solution including Social Desirability
This section will include an exploration of the five and three factor solutions conducted for
the SAPI clusters, including Social Desirability. Judging from the results of these factor
solutions, there appears to be a possible problem in terms of the conceptualisation and/or
scale construction of this instrument as the clusters do not load as expected.
The five factor solution conducted did not replicate as theoretically proposed. Factor 1 of the
five factor solution represented a mixture of Soft-Heartedness (Active Support, Empathy, I
help others to succeed), Facilitating (Encouraging Others, Guidance) and Relationship
Harmony (Harmony Maintenance) sub-clusters; while Factor 2 similarly included an
amalgamation of Soft-Heartedness (Hostility, Egoism) and Relationship Harmony (Harmony
Breach, Approachability) sub-clusters. Factor 3 included the sub-cluster Social Desirability
and the Integrity cluster and Factor 4 represented the Impression Management sub-cluster of
the Social Desirability cluster. Lastly Factor 5 formed the Social Desirability sub-cluster Selfdeception (see Table 3.7).
116
The three factor solution conducted revealed the existence of an overarching three factors;
Factor 1 was representative of Soft-Heartedness (Active Support, Empathy, I help others to
succeed), Relationship Harmony (Harmony Maintenance), Facilitating (Encouraging Others,
Guidance) and Integrity. Factor 2 included an amalgamation of Soft-Heartedness (Hostility,
Egoism) and Relationship Harmony (Harmony Breach, Approachability) sub-clusters. Factor
3 was representative of the Social Desirability cluster with the inclusion of all three subclusters- Social Desirability, Impression Management and Self-deception (see Table 3.8).
Factor 1 and 2 of these factor solutions reveal that different elements of the Soft-Heartedness
and Relationship Harmony clusters conglomerate to assess a similar theme. Nel (2008) stated
that Relationship Harmony and Soft-Heartedness are very closely related, but distinct. SoftHeartedness describes the interpersonal understanding and consideration of a person towards
others, while Relationship Harmony includes the building and maintaining of healthy and
constructive relationships with others (Nel, 2008). Accordingly, Factor 1 and 2 appear to be
tapping into both these definitions in opposite ways. The clusters therefore describe the
defining attributes of collectivism; which include defining oneself as an aspect of a group to
which one is interdependent, an overlap in personal and group goals such that social
behaviour is predicted by group norms, and considering relationships to be of the utmost
importance (Markus &Kitayama, 1998; Green, at al., 2005; Singelis,Triandis,
Bhawuk&Gelfand, 1995).Furthermore, the clusters Soft-Heartedness and Relationship
Harmony are said to correspond the strongest with the description and meaning of Ubuntu
(Nel, 2008).
Factor 1 can therefore be said to assess the collective element of personality, as it describes
an individual who seeks to maintain harmonious relationships assimilating in to the collective
group (Triandis, 1989; Markus &Kitayama, 1998); directly related to the concept of Ubuntu
(Nel, 2008). Factor 2, on the other hand, describes those individuals who do not seek
relationality or harmonious relationships, therefore assessing a more individualistic element
to the collective dimension of personality. Thus it can be said that these factors provide
evidence of collectivism in personality in South Africa, highlighted by an element of the
spirit of community dictated by the concept of Ubuntu. The remaining factors of these
solutions appear to assess a more individualistic element to this collective dimension of
117
personality in South Africa, related to an individual’s behaviour in terms of relationships with
others.
The ‘I help others to succeed’ sub-cluster of the Soft-Heartedness cluster appears to load
appropriately in these two solutions; loading together with the other Soft-Heartedness subclusters on Factor 1 as theorised. A surface level exploration of the Soft-Heartedness cluster
and the ‘I help others to succeed’ sub-cluster illustrates that this sub-cluster should include an
interpersonal understanding and consideration of a person towards others (Nel, 2008); as the
definition of Soft-Heartedness describes. However, a closer inspection of this sub-cluster
reveals that it appears rather to assess a more active element of caring for others similar to the
definition of Facilitating- guiding, leading, or advising others through life challenges and to
encouraging and them to improve themselves (Nel, 2008). Therefore it could be argued that
the sub-cluster ‘I help others to succeed’ should be incorporated into the Facilitating cluster
of the SAPI- an element of the Soft-Heartedness cluster (as argued in section 4.7.1.2. below);
however further research into this suggestion is required.
The five factor solution further reveals that the Social Desirability cluster may likewise
possess an issue with regard to its conceptualisation and/or scale construction; as the subclusters do not load onto one factor. The inclusion of the fifth factor reveals the presence of
some kind of Social Desirability factor; however the three sub-clusters- Social Desirability,
Impression Management and Self-deception- load onto three separate factors. The Social
Desirability cluster as mentioned previously includes items from the MCSDS and the BIDR;
thus assessing different components of this construct (Leite&Beretvas, 2005; Paulhus, 2002;
Uziel, 2010). The Social Desirability sub-cluster of the MCSDS loads with the Integrity
cluster in this solution, therefore suggesting that Integrity and Social Desirability may be
tapping into the same construct and should possibly be reconceptualised and combined as one
factor. This argument is in line with the discussion provided in the reliability section 4.5. of
this study in relation to the definitions provided for the composites of Social Desirability and
the Integrity cluster by Paulhus (2002) and Nel (2008) respectively; which dictate an overlap
in the conceptualisation of these clusters. Social Desirability aims to assess those behaviours
that are knowingly or unknowingly part of one’s true nature, which one can be honest or
dishonest about (this can be a conscious or unconscious choice) (Uziel, 2010). It would then
118
appear that the resultant honesty or dishonesty forms a component of the moral soundness of
the individual and their tendency to act in an honest, pure, and responsible manner; as the
Integrity cluster suggests (Nel, 2008).
A review of the literature on social desirability over the last four years indicates that although
these scales are largely adequate indicators of social desirability, their functionality has come
into question; particularly with regards to dimensionality (Leite&Beretvas, 2005; Li &Reb,
2009;Uziel, 2010). Despite its frequent use, researchers have criticised the MCSDS for its
ambiguous scale structure, its low reliability and its validity as a fake detector (see
Leite&Beretvas, 2005; McCrae & Costa, 1983; as cited in Uziel, 2010). Additionally, the
length of the scale and the outdated wording place the practicality of the scale in doubt
(Strober, 2001; as cited in Uziel, 2010). Furthermore, the social constructs of the original
MCSDS may not be as relevant to the social context in South Africa (Verardi et al., 2010).
Generally, a lack of clear dimensionality has also emerged for social desirability scales;
however most researchers in this field have adopted a two-dimensional factorial structure of
social desirability (Uziel, 2010). In particular, research by Wiggins (1964; as cited in Uziel,
2010) indicated the emergence of two major factors- named Alpha and Gamma; with
MCSDS found to have a high loading on the Gamma factor, providing possible evidence for
a one factor structure. Paulhus’ BIDR (2002) adopted a two largely orthogonal dimension
structure; impression management and self-deception. In this regard, the MCSDS was found
to be stronger marker of the impression management factor than any other factor; indicating a
possible one factor structure (Uziel, 2010). However, there is no clear evidence in research on
the MCSDS to support this fit of responses to one factor- need for approval (Leite&Beretvas,
2005).
Some authors, however, have proposed that MCSDS scores represent a two-factor model
(Leite&Beretvas, 2005). It was hypothesised that positively and negatively keyed MCSDS
items represent the Attribution factor and the Denial factor respectively; as discussed in the
reliability section 4.5. above (Ramanaiah, Schill& Leung, 1977; as cited in Leite&Beretvas,
2005). However such results need to be read with caution as the resultant evidence in fact did
not support a two-factor structure (Leite&Beretvas, 2005).A study conducted by Verardi et
al. (2010) found evidence of two factor- achievement and interpersonal relationship-
119
comparable to Paulhus’ (2002) self-deception and impression management respectively; thus
indicating the cross-cultural generalisability of the MCSDS. However, a review of item bias
indicated a limitation to this cross-cultural validity. Another study conducted by Leite and
Beretvas (2005) revealed that the theoretical basis of the MCSDS and the BIDR; the one- and
two- factor model fit was not supported. However, responses to the self-deception scale fit a
single factor model, leading Leite and Beretvas (2005) to suggest that the impression
management responses are likely multidimensional, as well as the responses to the MCSDS;
yet no clear evidence has emerged in this respect (Verardi et al., 2010).
The results of the five factor solution of the SAPI clusters indicate that the Social Desirability
cluster may be problematic; with particular reference to the Social Desirability sub-cluster
containing items from the MCSDS, as this sub-cluster loaded with the Integrity cluster in this
solution. In light of the research on the MCSDS it appears that in recent years this scale has
become increasingly problematic. It is particularly relevant to highlight that little evidence is
available for the need for approval factor structure measured in this scale (Leite&Beretvas,
2005); indicating that this scale may in fact be measuring an element of Integrity instead of
Social Desirability, as the five factor solution suggests, and may thus be better represented
under the Integrity cluster. However, further research is warranted in this regard.
Furthermore, research on the MCSDS has indicated that Social Desirability in this scale is
regarded as a personality trait more than a response bias (Verardi et al., 2010). Little support
has also been provided for a two factor model fit of the BIDR (Leite&Beretvas, 2005).
Alternatively, the Integrity cluster of the SAPI does not include any sub-clusters and does not
load on its own factor in these two solutions; thus placing the validity of this cluster as its
own factor in doubt. This indicates that the Integrity cluster consequently needs to be
rethought, reconceptualised and reconfigured.
In the five factor solution of this study the Impression Management and Self-deception subclusters of the Social Desirability cluster load onto different factors indicating that they may
be assessing separate constructs, rather than different elements of the same dimension.
Alternatively, this therefore may provide evidence of a two factor structure for the BIDR as
theorised or could suggest that these sub-clusters need to be reworked, within the South
African context, in order to assess Social Desirability as one factor. The self-deception
120
dimension is said to represent a fascinating indicator of social desirability; however its
function in detecting social desirability bias is said to be limited due to inability to obtain
conscious efforts at deception (Uziel, 2010). However the clear emergence of a Selfdeception factor in this solution provides evidence as to its function and adequacy as an
indicator of social desirability bias.
In the three factor solution, contrastingly, Integrity loaded separately to Social Desirability
and its sub-clusters; with Integrity loading onto Factor 1. This may indicate that the Integrity
cluster is not an adequate measurement of this dimension, as it does not load onto its own
factor. However, Factor 1 has been described as tapping into the collective dimension of
personality in South Africa related to an element of the concept of Ubuntu, thus confirming
that proposed by Nel (2008), who indicated that Soft-Heartedness, Relationship Harmony and
Integrity correspond the strongest with the description and meaning of Ubuntu (Nel, 2008). In
this solution, however, Integrity was found to cross-load with Factor 3- representative of the
Social Desirability cluster. This further emphasises the overlap between these two clusters;
thus indicating the need for the reconsideration and reform of these clusters. Furthermore it is
evident that Integrity is not a valid measure on its own and thus needs to be rethought. The
Social Desirability cluster, on the other hand, loaded correctly in this solution, with all three
sub-clusters loading onto one factor. This therefore suggests that the three factor solution
provides the best fit for the SAPI clusters, as the sub-clusters clearly load the closest to that
theorised by Nel (2008) and should therefore be considered by the developers of the SAPI.
4.7.1.2. Factor Solution excluding Social Desirability
This section will include an exploration of the four and two factor solutions conducted for the
SAPI clusters, excluding Social Desirability.
The four factor solution of the SAPI did not replicate as expected, with different sub-clusters
of different clusters loading onto each factor. Factor 1 of the four factor solution represented
a mixture of Soft-Heartedness (Active Support, Empathy), Facilitating (Encouraging Others,
Guidance) and Relationship Harmony (Harmony Maintenance) sub-clusters; while Factor 2
121
similarly included an amalgamation of Soft-Heartedness (Hostility, Egoism) and Relationship
Harmony (Harmony Breach, Approachability) sub-clusters. Factor 3 included the sub-cluster
‘I help others to succeed’; indicative of the Soft-Heartedness cluster and Factor 4 represented
the Integrity cluster (see Table 3.11).
The two factor solution conducted similarly revealed the existence of an overarching two
factors; Factor 1 was representative of Soft-Heartedness (Active Support, Empathy, I help
others to succeed), Relationship Harmony (Harmony Maintenance), Facilitating
(Encouraging Others, Guidance) and Integrity. Factor 2 included an amalgamation of SoftHeartedness (Hostility, Egoism) and Relationship Harmony (Harmony Breach,
Approachability) sub-clusters (see Table 3.12).
Factor 1 of these solutions -as discussed in section 4.7.1.1.above- is representative of
collectivism in personality in South Africa, in which individuals tend to see themselves as a
fabric of a social network (Li &Reb, 2009). This factor appears to be reflective of the relation
between human experiences and its definition of the self in a social and interpersonal context,
such that it measures an indigenous aspect of personality in South Africa related to
collectivism (Ho et al., 2001; Nel, 2008). Similarly, this collectivist dimension appears to
embrace an element of Ubuntu as dictated by Nel (2008). Factor 1, of both the solutions,
encompasses the Soft-Heartedness sub-clusters Empathy and Active Support that comprise of
aspects of personality related to the spirit of Ubuntu of a collectivist South African culture;
such as caring and loving, compassionate, considerate, satisfying others, community
involvement, generous, helpful and supportive, to solve problems of others, and heedful (Nel,
2008). Relationship Harmony comprises of some terms related to Ubuntu, such as having
good relations with others, maintaining constructive relations, and being peaceful, as well as
to be the peacekeeper in order to build and sustain good relationships with others- indicative
of collectivism (Nel, 2008).
The Facilitating sub-clusters- Encouraging Others and Guidance- appear to likewise assess an
element of collectivism. While it appears, at a surface level analysis, that Facilitating may
assess this dimension largely in terms of an understanding and consideration towards others
122
(indicative of Soft-Heartedness); the elements of the Facilitating cluster could arguably be
used as a means with which to facilitate the building and maintaining of relationships
(indicative of Relationship Harmony). Soft-Heartedness is broadly defined as the gentle side
of people caring about the welfare of others, either actively or passively (Louw, 2002; Nel,
2008). Facilitating, on the other hand, is specifically described as actively or passively caring
about the welfare of others through guiding, leading and encouraging others (Nel, 2008). This
therefore prompts me to suggest that theoretically, Facilitating may form part of an aspect of
Soft-Heartedness, particularly the Active Support element) and therefor may be included as
part of this cluster. However, further research is required to confirm this.
Factor 2 of these solutions, contrastingly, appears to be assessing a construct that almost
opposes the spirit of Ubuntu and collectivism or rather the aspects of oneself that dictate a
disregard of others and a relationship between them. This may highlight an important feature
of personality within South Africa that could be a result of the effects of acculturation (as
discussed in the next section), which could lead to a more individualistic orientation among
South Africans and particularly South African students (Mpofu, 2001). Individualism
characterises the self as egocentric, with a focus on independence, autonomy and personal
goals (Green et al., 2005). While the collective construction of personality refers to
relationality; individualistic constructions view and describe groups in terms of individual
identities submerged within a social identity (Ellemers, Spears &Doosje, 2002;Markus
&Kitayama, 1998). Thus it appears that Factor 2 may describe an individualistic expression
of this collective dimension of personality in South Africa.
The Individualism-Collectivism dimension, highlighted in Factor 1 and 2, can be further
distinguished in that a distinction can be made between vertical and horizontal individualism
and collectivism. This distinction is due to the degree to which equality among individuals is
emphasised (horizontal) or the degree of hierarchy (vertical) in which individuals differ in
status (Singelis et al., 1995; Triandis, 2001; Green et al., 2005). Horizontal collectivism is
described as a cultural pattern in which the individual sees the self as an aspect of the ingroup, characterised by equality and interdependence. Vertical collectivism is a cultural
pattern in which the individual sees the self as part of the in-group; however inequalities exist
in which some members have more status than others (Singelis et al., 1995; Green et al.,
123
2005). Horizontal individualism is a cultural pattern where the individual is independent and
autonomous, but is equal in status to others. Vertical individualism describes a cultural
pattern, in which the individual sees an autonomous self; however differences exist in
equality and status among individuals as a result of competition (Singelis et al., 1995; Green
et al., 2005).
Furthermore, Factor 1 and 2 of this solution can be further distinguished in terms of these
four cultural patterns. Factor 1 of the SAPI factor solutions has been described as an element
of the collectivism dimension of personality in South Africa; likened somewhat to the
essence of Ubuntu. This therefore prompts me suggest that this factor fits in with the
horizontal dimension of the Individualism-Collectivism distinction, characterised by equality
and being a member of the in-group (Singelis et al., 1995; Green et al., 2005). Factor 1- the
collectivistic dimension- and horizontal collectivism describe the interdependence of the
individual to the in-group; while horizontal individualism focuses on equality emphasising
relationships with others and societal harmony (Markus &Kitayama, 1998; Singelis et al.,
1995; Green et al., 2005). The different cultural groups of South Africa have been
distinguished along individualistic and collectivistic lines, indicating cultural differences in
the emphasis on societal values (Nel, 2008; Meiring, 2006). However, the emergence of
Factor 1 indicates the extension of the collectivistic dimension of personality and the concept
of community, indicative of Ubuntu, to include individualistic and collectivistic expressions,
highlighted by the horizontal dimension. Conversely, Factor 2 of the SAPI factor solutions
has been described as the antithesis of the spirit of community and collectivism; likened to
the vertical dimension of the Individualism-Collectivism distinction. Factor 2 incorporates
individualistic elements that dictate a disregard for others and societal harmony; while the
vertical dimension describes as inequality in status indicative of an individualist orientation
(Markus &Kitayama, 1998; Singelis et al., 1995; Green et al., 2005). Vertical collectivism, in
terms of Factor 2, highlights an individualistic expression within one’s membership to the ingroup; contrastingly vertical individualism emphasises an autonomous being different to the
group as a result of competition- characterised by the sub-clusters of Factor 2 such as
Hostility and Egoism.
124
Factor 3, of the four factor solution, includes the ‘I help others to succeed’ sub-cluster of the
Soft-Heartedness cluster. This sub-cluster loads onto its own factor in this solution, as
opposed to onto Factor 1 like in the two, three and five factor solutions; suggesting that
further research is required in order to investigate whether this sub-cluster deserves its own
factor. At surface level it appears to assess something akin to Facilitating as part of SoftHeartedness, discussed in the previous solutions. This is further emphasised in the four factor
solution, as the Facilitating sub-clusters-Encouraging Others and Guidance- cross loaded
Factor 3; indicating that further research into the inclusion of ‘I help others to succeed’ into
the Facilitating cluster is necessary. Furthermore, this sub-cluster comprises of one item such
that it is deemed to be unreliable (Carmines & Zeller, 1979); as a result further research into
this sub-cluster is required.
Factor 4 of the four factor solution indicates that Integrity loaded alone, without the inclusion
of Social Desirability, in the solution. Previous solutions that included Social Desirability
resulted in Integrity assessing something akin to the collectivism dimension that
encompassed an element of Ubuntu, as well as assessing a construct similar to the Social
Desirability sub-cluster. However, this solution indicates that without the presence of Social
Desirability, Integrity was found to be its own factor. Integrity was found in the four factor
solution to cross load onto Factor 1- the collectivism dimension related to an element of
Ubuntu; while furthermore Egoism cross-loaded onto the Integrity factor. This therefore
suggests that the Integrity cluster needs refinement.
It appears from these solutions that the four factor solution provides the better fit, as Integrity
emerges as a separate factor as theorised by the SAPI developers. However, the two factor
solution provides a clear distinction in terms of the Individualism-Collectivism and VerticalHorizontal distinction. The two factor solution clearly indicates that the four collective
dimensions of the SAPI can be divided in two factors, in which Factor 1 describes a
collective horizontal dimension of personality; while Factor 2 describes an individualistic
expression of the collective vertical dimension of personality in South Africa.
125
4.7.2. Item Level: SAPI
Construct validity of the SAPI was conducted at item level using exploratory factor analysis
with varimax rotation. As indicated previously, theoretically the SAPI encompasses four
broad collective dimension clusters- namely Soft-Heartedness, Facilitating, Integrity
andRelationship Harmony- and a fifth cluster Social Desirability; consequently a five factor
solution was conducted on the SAPI items. The five proposed clusters include fourteen subclusters that constitute these clusters; accordingly a fourteen factor solution was explored at
item level. Statistically the eigenvalues of the SAPI item level analyses indicated the
retention of seventy five factors; this is however not tenable, therefore only a five and
fourteen factor solution were explored. Results of these two solutions indicate that the items
did not load as expected.
The order of the factors in this solution may also provide some important and interesting
insight into the collective dimension of personality in South Africa; however further research
is warranted. The order of the factors in the five and fourteen factor solutions appeared to be
somewhat different, in terms of the themes that emerged. An overarching four themes
emerged at item level- these included a collective dimension indicative of harmonious
relationships, facilitating of the building and maintenance of these relationships, an
individualistic expression of the collectivistic dimension related to the individual’s behaviour
in terms of relationships with others, and lastly an element of Integrity and Social
Desirability. The total variance explained for the five and fourteen factor solutions were
32.52% and 43.73% respectively; thus indicating that the additional nine factors in the
fourteen factor solution are not really adding anything substantial to the five factors. When
considering the variance explained by each factor it is evident that the first four factors
explain the most variance in these solutions, with Factor 1 explaining the most variance with
19.14%, followed by Factor 2 with 6.29%, then Factor 3 with 2.95% and Factor 4 with
2.35%.
126
When considering these figures, it seems plausible to suggest that Factor 1 and 2 may be
contributing the most to an understanding of personality, such that they may be the most
important determinants of the collective dimension of personality in South Africa.
130
Interestingly, Factors 1 and 2 emerge as different factors for the five and fourteen factor
solution at item level. Factor 1 and 2 for the five factor solution are indicative of the
collective dimension of personality in South Africa in terms of one’s care for others,
interpersonal roles and the facilitating and maintenance of their relationships with others;
while Factor 1 and 2 of the fourteen factor solution describe the facilitating element in one’s
relationships of the collective dimension and a more egocentric orientation with regards to
one’s relationships with others indicative of a more individualistic orientation, respectively
(Green et al., 2005; Nel, 2008).
South Africans may function in collectivistic and/or individualistic environments and thus
they adopt the necessary qualities of each environment depending on the situation (Nel,
2008). Nonetheless, all South Africans have been shown to possess a strong inclination to be
sensitive towards others and to build and maintain a community; thus revealing a collective
dimension (Nel, 2008).It is clear from these item level analyses that the SAPI is in fact
assessing the collective dimension of personality in South Africa; however the theoretical
foundation of the SAPI needs to be revised in order to include the salient aspects of the South
African personality in the ways that they empirically emerge, further research is still required.
Protocol from the developers of the SAPI showed no indication of reverse scoring in the
instrument. However, the item level analyses revealed numerous negative factor loadings as
well as ambiguous items, indicating that the developers need to consider reverse scoring and
the wording of each item in order to combat this.
4.7.2.1.Fiveand Fourteen Factor Solution
An exploratory factor analysis conducted at item level for the SAPI using a five factor
solution indicated the existence of only four factors. This analysis at item level provided
different results to those obtained at scale level, with items across clusters and sub-clusters
loading onto each factor. At item level, Factor 1 was largely representative of SoftHeartedness and Relationship Harmony items, Factor 2 included mostly Facilitating items,
Factor 3 once again represented mostly Soft-Heartedness and Relationship Harmony items, as
131
well as a few Integrity and Social Desirability items, and lastly Factor 4 included a
combination of Integrity and Social Desirability items.
A fourteen factor varimax solution was conducted at item level for the SAPI sub-clusters; the
fourteen factors do not load as expected, with only thirteen of the fourteen factors obtaining
factor loadings, as well as an almost haphazard loading of items across clusters and subclusters that are somewhat difficult to discern. An exploration of these sub-clusters reveals
that an overarching general four themes tend to emerge from all the sub-clusters, much like
the five factor solution discussed in section 4.7.1.1. above. Factor 1 of this solution largely
incorporated the Facilitating sub-clusters- Guidance and Encouraging Others, as well as the
Active Support sub-cluster of Soft-Heartedness. Factor 2 represented mostly SoftHeartedness (Hostility, Egoism) and Relationship Harmony (Approachability, Harmony
Breach) items, as well as a few Integrity and Social Desirability items; while Factor 3 was
found to represent largely Soft-Heartedness (Empathy, Active Support) and Relationship
Harmony (Harmony Maintenance) sub-clusters. At best Factor 4 represented mostly Integrity
items and Factor 8 included Integrity and Impression Management items. The resulting
factors, however, include largely disorganised individual item loadings that were difficult to
decipher.
In accordance with the factor analysis conducted at scale level, Factor 1 1(of the five factor
solution) and Factor 3 (of the fourteen factor solution) of the item level analyses
includedSoft-Heartednesssub-clusters Active Support and Empathy, Relationship Harmony
sub-cluster Harmony Maintenance, and Facilitating sub-cluster Encouraging Others.This
factor therefore describes arelationality emphasised in collectivistic culturesand the
horizontal dimension (Markus & Kitayama, 1998), which dictates a care toward others and
allows for the maintenance of supportive relationships with others as indicated by the concept
of Ubuntu (Nel, 2008). Factor 1and 3 are largely reflective of Factor 1 at scale level,
however,at item level they seem to deviate from the scale level analysis in terms of its lack of
incorporation of most of the Facilitating cluster.
1
The five factor solution results will be placed in italics for the item level analyses.
132
Factor 1and 3 also appear to include a haphazard loading of items from other clusters and
sub-clusters of the SAPI.Items 41 2of the Hostility sub-clusterand 205 3 of the Egoism subclusterload onto Factor 1 and 3; thus indicating that they are tapping into the theme of this
factor more so than the Soft-Heartednesssub-clusters in which they originate.On inspection of
the items it appears that in order for these items to load effectively onto their respective subclusters appropriately (i.e. Hostility and Egoism) they need to be reverse scored. Item 200 4 of
the Approachability sub-cluster conversely loaded negatively onto Factor 1for the five factor
solution and thus needs to be reverse scored in order to affectively measure the construct that
this factor is aiming to measure.The Approachability items- 136, 188- of both solutions
reflect an individual who is open and available in terms of relationships with others; thus
reflecting a desire to preserve harmony in one’s relationships indicative of the collectivistic
dimension of personality (Markus & Kitayama, 1998; Singelis et al., 1995). Similarlythe
Integrity item 235 5, and the Encouraging Others items 38, 91 of both solutions, as well as
item 192of the five factor solution, reveal individual characteristics that reflect an interest in
the well-being of others and the maintenance of successful relationships with them- indicative
of the spirit of community and collectivism (Nel, 2008).
In contrast to the scale level analysis, the Facilitating items- Guidanceand Encouraging
Others - loaded separately onto Factor 1 and Factor 2.This indicates that at item level, the
Facilitating cluster largely forms its own factor.However, at scale level the Facilitating cluster
formed part of an element of the collective dimension of personality in South Africa, along
with the concept of Ubuntu (Factor 1); whereas at item level this cluster appears to assess an
aspect of the collective dimension of personality separate to this factor. These item level
analyses therefore suggest that guiding, leading and encouraging others through life’s
challenges is a separate construct from caring for others and maintaining relationships with
them (Nel, 2008). In section 4.7.1., contrastingly, it was suggested thatFacilitating may form
part of the Soft-Heartedness cluster. Thus further research is required in order to determine if
Facilitating is in fact its own factor or if it forms part of the Soft-Heartedness cluster.
2
‘I care for others like for my own family.’
‘I share what I have with others.’
4
‘I am better than others.’
5
‘I give everyone a chance.’
3
133
In addition to these items, this factor includeditems from the Soft-Heartednesssub-cluster
Active Support and from the Relationship Harmony sub-cluster Harmony Maintenance. On
closer inspection of these items they appear to assess aspects of personality that relate to
being helpful, supportive, protective, involved in the community and solving the problems of
others.The sub-cluster‘I help others to succeed’ (item 275) appears, as discussed in the scale
analysis above, to be incorporated into the Facilitating cluster of this factor, therefore
somewhat providing evidence for the inclusion of this sub-cluster into the Facilitating
cluster.Item 55 (‘I criticise people when I do not understand them’) from the Hostilitysubcluster and item 240 (‘I provide for those who are in need’) from the Empathy sub-cluster
loaded onto Factor 1 and 2 and are thus in line with the general theme of this factor that talks
to the spirit of community that encourages, helps, leads and uplifts others in terms of the
difficulties that they encounter in life or in their relationshipsthrough on-going support,
involvement and protection (Nel, 2008).This emphasises an element of collectivism by which
individuals evaluate themselves in terms of others,their interpersonal roles and the quality of
their relationships (Green et al., 2005).This factor could furthermore be said to represent an
element of Ubuntu as it reflects the spirit of unity in a community, placing an emphasis on
collective collaboration (Nel, 2008). However, further research is required in this regard.
Factor 2 and Factor 3 were an amalgamation of the Soft-Heartednesssub-clusters Hostility
and Egoism, and the Relationship Harmony sub-clusters Harmony Breach and
Approachability, as well asitems from the Integrity cluster and from the Social Desirability
cluster.Factor 2 and 3 appear to be somewhat similar to Factor 2 of the scale level analysesrepresentative of a neglect of others and one’s relationships; thus reflecting a more
individualistic orientation as indicated by the vertical dimension as opposed to an emphasis
on interdependence, collective aims and harmonious relationships (Green et al., 2005). The
item level analysis, howeverin contrast, included numerous Integrity and Social Desirability
loadings; however on examination of the Social Desirability items such as items 140 6and
224 7, as well as 252 8 of the five factor solution, it is clear that these items tap into the nature
of this facto. This factor describes an individual who is jealous, aggressive, self-centred,
unapproachable, intolerant, conflict-seeking and possesses a lack of morality (Nel,
6
‘I gossip.’
‘I let someone else be punished for my wrongdoings.’
8
‘I am jealous of others with good fortune.’
7
134
2008).Item 14- ‘When I hear people talking privately, I avoid listening’- formed part of the
Impression Management sub-cluster and loads negatively onto Factor 2 and 3; in order to
appropriately assess the theme of this factor this item needs to be reverse scored. Item 54-‘I
give people directions in life’- of the Guidance sub-cluster, contrastingly, loaded positively
onto Factor 2 and 3; however it should have loaded negatively in order to tap into the theme
of the factor. Therefore this item needs to be reassessed and reworked.
Factor 4 of the five factor solution incorporated four Integrity items and four Social
Desirability items. The Integrity items included item numbers 83, 148, 195 and 260- an
examination of these items revealed the assessment of a moral component related to one’s
principles and moral beliefs. Two Social Desirability items- 196 and 228- were found to talk
to an individual doing things one’s own way; while the two Self-deception items- 51 and 89included an element of stubbornness and confidence in one’s judgments. This factor can thus
be said to be reflective of the strength of one’s morals,truthfulness and the degree to which
the individual may succumb to social desirability (Nel, 2008; Uziel, 2010).This factor can
therefore be compared to Factor 3 of the five factor solution at scale level, as it similarly
appears that Integrity and Social Desirability may be tapping into the same construct such
that they may be combined into one factor. Further research is required in this regard.
Factor 8 of the fourteen factor solution, contrastingly, included three negative Integrity item
loadings; item 2 (‘I am loyal to others’), item 8 (‘I tell the truth’), item 129 (‘I keep my
promises’). This factor also included the positive Integrity loading from item 222 (‘I make up
false stories’), as well as the positive Impression Management items 69 (‘I sometimes tell lies
if I have to’) and item 206 (‘I have sometimes dropped litter on the street’). Factor 8 thus
talks to a loss of honesty, truthfulness and integrity in terms of relationships; such that the
positive loadings on this factor (items 206, 222 and 69) need to be reverse scored in order to
be appropriate for this factor. This factor can therefore almost be described as the reverse of
Factor 4 of the five factor solution at item level, discussed above, as it talks to one’s moral
fibre in respect to others. This provides further evidence that Integrity and Social Desirability
overlap in terms of their definition and assessment of these constructs.
135
Factor 4 of the fourteenfactor solution included various loadings from four different subclusters of the SAPI. An examination of the items present in this factor reveals the Integrity
items- 122, 195 and 212- to be indicative of doing the right thing and taking care of one’s
problems (Nel, 2008). Factor 4 also includes item 226 (‘I earn other people’s respect’) from
the Guidance sub-cluster, item 231 (‘I speak politely to others’) from the Harmony
Maintenance sub-cluster and lastly item 164 (‘I speak on behalf of others when they cannot
defend themselves’) from the Active Support sub-cluster. This factor is thus difficult to
determine, however it appears to delineate a moral component emerging from one’s society;
therefore incorporating elements of honesty, responsibility and moral consciousness (Nel,
2008). Therefore these items need to be reconceptualised such that it is incorporated into a
particular sub-cluster or the items of these sub-clusters need to be reworked in order to tap
into the various constructs that they propose.
Factor 5 of the fourteen factor solution included four factor loadings from three different subclusters across SAPI clusters.This factor encompassed two Egoism items- 125 (‘I want more
than what is enough’) and 178 (‘I like having everything my way’). In addition, item 196- ‘I
do things my way’- of the Social Desirability sub-cluster and the Integrity item 117 (‘I favour
some people above others’) were also incorporated into Factor 5. This factor thus illustrates
an egocentric and selfish aspect of personality that inhibits a concern for others and one’s
relationships (Nel, 2008). Theoretically, this factor appears to assess an element of Factor 2
and 3 at item level; therefore further research is required in order to confirm or dispute this.
Factor 6 of the fourteen factor solution experienced nine loadings across all five clusters of
the SAPI; as a result this factor is difficult to discern. The three Empathy items- 250, 265 and
273 that loaded onto this factor talk to the caring and loving nature of an individual towards
others. The two Egoism items- 263 (‘I only care about my things’) and 264 (‘I dislike it when
others get ahead in life’) - loaded negatively onto Factor 6, indicating that these items need to
be reverse scored in order to accurately assess the caring and loving nature of an individual
described by this factor. Another four items from different sub-clusters load onto Factor 6;
these included item 274 9 of Encouraging Others, item 253 10 of Harmony Maintenance, item
9
‘I like seeing people progress in life.’
136
252 11 of Social Desirability and item 251 12 of Integrity. Items 252 and 251 loaded negatively
onto this factor and therefore need to be reverse scored in order to fit into the theme of Factor
6. This factor appears to be the antithesis of Factor 5 of the fourteen factor solution, as it
speaks to an element of collectivism and the concept of Ubuntu and the need to maintain
harmonious relationships in order to be part of the collective (Triandis, 1989; Markus &
Kitayama, 1998). As a result, it compels me to suggest that Factor 6 is in fact assessing
ancollectivistic aspect of personality dictated by Factor 1 and 3 at item level and therefore
may be incorporated there. In order to do so this factor needs to be reworked and further
research is required.
Factor 7 of the fourteen factor solution does not seem to be representative of any particular
sub-cluster as it included five loadings from three sub-clusters across three clusters. This
factor included the positive Guidance items 167 13 and 175 14, the negative Integrity items 16
(‘I hide from others who I really am’) and 174 (‘I give others the wrong impression of
myself’); as well as the negative Hostility item 118 (‘I find it difficult to confide in others’).
Factor 7 therefore seems to measure the degree to which individuals are morally sound and
reflect such qualities in their relationships with others (Nel, 2008). Items 16, 174 and 118
therefore need to be reverse scored in order to effectively measure what this factor is
intending to measure. It appears, on closer inspection, that this factor and Factor 4 of the
fourteen factor solution may be assessing different elements of the same construct in relation
to one’s individual demeanour or characteristics inherent in relationships with others and
therefore could be incorporated there. Further research for this proposition is required.
Factor 9 of the fourteen factor solution included the Approachability item 70- ‘I believe that
there is nothing to learn from other people’; reflective of arrogance and pride that prevents
relationships with others. However, this item only loaded moderately onto Factor 9 and
therefore may need to be rephrased or reworked to fit elsewhere. At face value it appears that
10
‘I listen patiently when others want to share their feelings.’
‘I am jealous of others with good fortune.’
12
‘I cheat.’
13
‘I am a person others love to be around.’
14
‘I attract the attention of people around me in a positive way.’
11
137
this item should in fact load onto Factor 2 and 3 at scale level, therefore more research is
necessary in this respect.
Factor 10 of the fourteen factor solution incorporated a loading from the Harmony
Maintenance item 22- ‘I want to sort things out in a peaceful manner’- and the Harmony
Breach item 130- ‘I avoid arguments’. These items are therefore indicative of the means by
which one solves their difficulties in relationships with others, which appears to be a
construct that has already been alluded to previously. Therefore these items may be reworded
or reconceptualised to fit elsewhere; however it seems that they could form part of Factor 1
and 3 at item level- indicative of a collectivistic orientation in personality.
Factor 11 of the fourteen factor solution comprised of three items; item 185 (‘I always obey
laws, even if I am unlikely to get caught’) from the Impression Management sub-cluster, item
7 (‘I am suspicious of what people close to me do’) from the Hostility sub-cluster and item 36
(‘I keep my things for myself’) from the Egoism sub-cluster. This factor is difficult to
decipher however it appears to incorporate aspects of individuals who are untrusting of
others. These three items need to be rephrased in order to effectively tap into their respective
sub-clusters; however at face value they seem to be tapping into Factor 2 and 3 at item level
that prevents the building and maintenance of relationships with others (Nel, 2008).
Factor 12 of the fourteen factor solution was found to include the Guidance item 93- ‘I am a
good leader’- thus this factor taps into an element of leadership with a moderate loading. This
item therefore appears to assess an element of Factor 1 and 2 at item level and consequently
needs to be reworked to be included in this largely Facilitating factor.
Lastly, Factor 13 of the fourteen factor solution included two Harmony Maintenance items 4
(‘I accept change’) and 193 (‘I apologise if I have made a mistake’) - thus delineating
flexibility, constructiveness and cooperation in relationships with others. These items,
tentatively, appear to assess an element of collectivism- as described by Factor1 and 3- that
138
deals with the maintenance of relationships. Therefore these items need to be rephrased in
order to be incorporated appropriately.
The item level solutions revealed a large number of items of the SAPI solution that did not
load on any factor 15. The 5 factor solution indicated a total of 96 of 290 items that did not
load onto any of the five suggested factors- particularly for the Integrity and Social
Desirability items.A total of 90 of 290 items did not load onto any of the fourteen suggested
factors for the fourteen factor solution- particularly for the Empathy, Active Support,
Integrity, Harmony Maintenance, Social Desirability and Self-deception sub-clusters. The
number of missing items in the item level analysis indicates that the inclusion of fourteen
sub-clusters may not be necessary. Thus these SAPI items need to be rewritten, reassigned or
reduced. Further research is required in this regard.
4.8. Construct Validity: CPAI-2
4.8.1.Scale Level
Construct validity of the CPAI-2 Interpersonal Relatedness collective dimension was
conducted at scale level using exploratory factor analysis with varimax rotation. Theoretically
the CPAI-2 postulates the existence of one factor- Interpersonal Relatedness- capturing six
scales; namely Traditionalism vs Modernity, Ren Qing, Social Sensitivity, Discipline,
Harmony and Thrift vs Extravagance.As a result a six factor solution was explored. The
resulting scree plot and eigenvalues, however, indicated the retention of two factors, therefore
a two factor solution was also explored.The total variance explained by the six and two factor
solutions was 100%and 58.78% respectively. These figures suggest that the six factor
solution in fact offers more insight into the collective dimension of personality assessed by
the CPAI-2 in the South African context than a two factor solution. This suggests that each
scale of the Interpersonal Relatedness factor may be adding something valuable to the
collective dimension.
15
68items did not load on the five or fourteen factor solution, these items include; 11, 18, 19, 20, 26, 27, 28,
30, 35, 37,40, 43, 44, 45, 46, 49, 50, 53, 56, 58, 59, 75, 77, 84, 85, 98, 103, 104, 107, 109, 112, 113, 114, 115,
121, 123, 124, 144, 153, 168, 171, 173, 179, 186, 189, 191, 197, 201, 204, 229, 234, 238, 242, 243, 248, 254,
256, 257, 258, 259, 269, 271, 277, 278, 279, 280, 282, 284
139
4.8.1.1. Six factor solution
An exploratory factor analysis with varimax rotation was conducted with the extraction of six
scales for the Interpersonal Relatedness factor in order to determine if the model proposed by
Cheung et al. (2008) would hold in a South African sample. Results of this solution revealed
the clear existence of all six scales in a South African sample; with Factor 1 found to consist
of the Harmony scale, Factor 2- the Traditionalism vs Modernity scale, Factor 3- the Thrift vs
Extravagance scale, Factor 4- the Ren Qing scale, Factor 5- the Discipline scale and finally
Factor 6- the Social Sensitivity scale (see Table 3.17). These results indicate that the
importance of social values and relationships- as informed by the Interpersonal Relatedness
dimension (Cheung et al., 2001)–was replicated in a South African context. Consequently it
may be concluded that these various aspects of “interdependence and reciprocity in
traditional social relationships” (Cheung et al., 2001, p. 425) are not only specific to Chinese
culture; but are rather representative of collectivist cultures evaluate themselves in terms of
others, their interpersonal roles and the quality of their relationships that they seek to
maintain (Green et al., 2005;Markus & Kitayama, 1998). Furthermore it appears as if this etic
measure is in fact contributing to an understanding of the indigenous collective dimension of
personality in South Africa.The existence of the Interpersonal Relatedness factor in this
study, in addition, provides evidence of the limitation of the FFM in describing personality;
however this is in contradiction to Laher (2010).
It is interesting to note that the Thrift vs Extravagance scale was found to exist in a South
African sample in this study. Previous research has indicated difficulties in replicating this
scale; with Thrift vs Extravagance not loading above .40 in a six factor solution in the
normative sample (Cheung et al., 2008). Similarly the Thrift vs Extravagancescale did not
load in a four or five factor solution of the CPAI-2 factors in a South African sample of
university students (Laher, 2010). This research suggests the problematic construction of the
Thrift vs Extravagance scale in measuring the collective dimension of personality in Chinese
culture or in a South African sample. When considering what this scale aims to measureone’s
carefulness in spending (Cheung et al., 2001) – it seems plausible to suggest that it may bea
cultural phenomenon, in which differences occur as a result of how these personality traits
are expressed in individualist and collectivist cultures (Allik & McCrae, 2004). Alternatively,
a problem could exist in the scale construction and makeup of the items of this scale.
140
The reliability of the Thrift vs Extravagance scale was found to be the most problematic
across all three studies- the current study, the normative sample and Laher’s (2010) studywith poor coefficients across all three studies, as indicated in section 4.6. The reliability was
found to be the lowest for this study (.36), followed directly by Laher’s (2010) study (.37),
with the high reliability found for the normative Chinese sample (.57).However, this solution
of the current study appears to provide evidence of the Thrift vs Extravagance scale as a
collective dimension of personality in the South African context. Additionally, this was found
to explain the third most variance with 14.46%, indicating that it may be adding something
significant to the collective dimension of personality in South Africa. Further research is thus
required in terms of this scale.
Similarly, the scales Ren Qing and Social Sensitivity did not achieve factor loadings of above
.40 in the normative sample (Cheung et al., 2008). This stands in contrast to the results of this
study that found both these factors to load strongly on their own factor; howeverin Laher’s
study these two scales did not load on the Interpersonal Relatedness factor in the four factor
solution, while just Social Sensitivity did not load on the Interpersonal Relatedness factor in
the five factor solution. Ren Qing or Relationship Orientation theoretically measures the
degree to which individuals adhere to cultural norms; while Social Sensitivity measures the
extent to which individuals are sensitive to others (Cheung, 2006; Cheung et al., 1996;
Cheung et al., 2001). In consideration of these definitions, as well as outer kingliness (or
one’s love for people through propriety; Cheung et al., 2008) it appears that these scales may
be comparable to the concept of Ubuntu- in that itdescribes a person as a person through
others and the community (Louw, 2002; Nel, 2008).Support of this is provided by Nel (2008)
who theoretically found an arbitrary relationship between Ren Qing and Relationship
Harmony of the SAPI; similarly Social Sensitivity was found to incorporate many facets
inherent in the Relationship Harmony cluster.
Relationship Harmony, as noted earlier, is considered to represent an element of the concept
of Ubuntu and therefore in turn Ren Qing and Social Sensitivity (Nel, 2008). In this respect,
the replication of these two scales in a South African sample makes sense in that it provides
141
evidence for the collective dimension of personality in South Africa. However, in this regard
it is interesting to note that Ren Qing and Social Sensitivity explain less variance than a
number of other Interpersonal Relatedness scales, with 10.20% and 7.59% respectively.
Consequently the order of factor loadings in this solution may provide some interesting
information as to the collective dimension of personality in South Africa and should be
researched further.
Another interesting finding provided by this solution indicated Traditionalism vs Modernity
vs Modernity as a collective dimension of personality in South Africa, explaining a total
22.14% of the variance.Traditionalism vs Modernitycan be describedas the degree of one’s
individual modernisation; thus exploring one’s attitude toward Chinese beliefs and values
(Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001). In a society such as South Africa
ridden with cultural diversity it seems only fitting that one should protect their customs(as
high scores of this scale suggest) particularly when considering the resulting
injusticesemerging from a history of Apartheid. However, on the other hand, it would seem
that Apartheid would cause these individuals to in turn advocate democracy and personal
freedom (as low scores of this scale suggest) (Cheung, 2006; Cheung et al., 1996; Cheung et
al., 2001).In Laher (2010) Traditionalism vs Modernity was found to load as expected with
the Interpersonal Relatedness factor for the four and five factor solution; however these
loadings were small to moderate.Acculturation may also play a role in the diffusion of one’s
customs and beliefs due to the inclusion of other cultural influencesand a move towards
western individualism (Eaton & Louw, 2002; Heuchert et al., 2000; Mpofu, 2001).
A study conducted by Laher (2010) in a South African sample using the CPAI-2 concluded
that the Interpersonal Relatedness factor is in fact representative of the
individualism/collectivism distinction. In terms of the results of this study, I would argue that
the Interpersonal Relatedness factor encompasses a collective dimension of personality that
may be expressed in a collectivistic or individualistic manner. It is therefore plausible to
suggest that the scales Ren Qing, Social Sensitivity and Traditionalism vs Modernityrepresent
a collective expression of the collective dimension of personality in their inclusion of an
interpersonal and relational element similar to the horizontal dimension; while Discipline,
Harmony and Thrift vs Extravagance represent an individual expression of the collective
142
dimension of personality in that these scales describe an individual expression that influences
one’s social demeanoursimilar to the vertical dimension (Markus & Kitayama, 1998).
Interpersonal Relatedness is therefore about achieving equilibrium through self-discipline and
one’s love and care for others (Cheung et al., 2008).In other words, a balance between one’s
individual and collective expressions of social-relatedness- or the collective dimension of
personality.This solution can therefore be said to provide evidence of the collective
dimension of personality in South Africa; however further studies are required.
The total variance explained by Factor 1- Harmony is 36.65% and Factor 2- Discipline is
8.95% in this solution. This may therefore indicate that these two factors may be the most
important determinants of the collective dimension of personality in South Africa; however
further research is required.Harmony is described to be the degree to which an individual is
content and possesses inner harmony and peace of mind (Cheung, 2006; Cheung et al., 1996;
Cheung et al., 2001). Although this dimension appears to express an individualistic element,
it suffices to say that such inner harmony would be imperative in one’s relationship harmony
with others. It would furthermore be plausible to suggest that this Harmony scale may reflect
an element of the concept of Ubuntu in South Africa. Nel (2008) provides support for this
suggesting an arbitrary relationship between Harmony and the Relationship Harmony cluster
of the SAPI- a composition of Ubuntu. The Discipline scale appears to represent inner
sageliness or rather self-discipline and personal cultivation inherent in the manifestation of
social values of the collective dimension of one’s personality. This element of personality
coupled with one’s love and care toward others is the manifestation of the Interpersonal
Relatedness factor, as evidenced by this solution in a South African sample (Cheung et al.,
2008).
4.8.1.2.Two Factor Solution
An exploratory factor analysis with varimax rotation was conducted with the extraction of
two factors as the empirical evidence suggested.Factor 1 includedthe scales Social Sensitivity
143
and Harmony; while Factor 2 consisted of the scalesTraditionalism vs Modernity, Discipline
and Ren Qing (see Table 3.18).The resultant two factor theory presented in this solution,
however contradicts the two factor theory suggested in the six factor solution above. The
scales of Interpersonal Relatedness do not appear to load as a collectivistic and individualistic
expression of the collective dimension of personality in South Africa as suggested. Therefore
two possible theories emerge for the loadings of the scales of Interpersonal Relatedness;
further research into this is required.
Social sensitivity describes an individual who possess qualities such asempathy,
understanding, willingness to listen and communicate, and those who are easily
approachable.Harmony dictates an individual who is even-tempered and peaceful(Cheung,
2006; Cheung et al., 2001; Cheung et al., 2008). Factor 1 thus incorporates the qualities of an
individual that dictates a concern for others and the ability to build and maintain
relationships. This could be equitable to the Relationship Harmony cluster of the SAPI, and
possibly even elements of the Soft-Heartedness cluster, that describe the concept of Ubuntu
in personality in South Africa (Nel, 2008).Nel (2008) further provides support for the
relationship between Social Sensitivity and Harmony of the CPAI-2 and the Relationship
Harmony cluster of the SAPI. This factor, therefore, includes both an individual (Harmony)
and a collective (Social Sensitivity) expression of the collective dimension of personality
related to one’s ability to obtain meaning relationships with others, as dictated by collectivist
cultures (Triandis, 1989; Markus & Kitayama, 1998).
Factor 2 of this solution likewise represents bothan individual (Discipline) and a collective
(Traditionalism vs Modernity and Ren Qing) expression of the collective dimension of
personality; however this factor appears to assess an element of Chinese Tradition in relation
to one’s adherence of cultural norms, beliefs and traditions (Cheung, 2006; Cheung et al.,
2001; Cheung et al., 2008). However the existence of this factor in a South African sample as
indicated by this solution reveals that this factor in fact represents an interdependent
communal adherence to norms and traditions as characteristic of African and Asian cultures
(Laher, 2010).The existence of this factor in a South African sample provides further
evidence of the universality of this collective dimension of personality.
144
The Ren Qing scale cross-loaded onto both factors in this solution, with a slightly higher
loading on Factor 2; however this can be expected as elements of personality are usually
related and are often found to overlap. With particular attention to the Ren Qing scale and its
definition it is clear that this scale taps into the theme of both Factor 1 and Factor 2
concerning effective relationships with others and the adherence to cultural norms
respectively; indicative of Chinese culture (Cheung et al., 2006).Nel (2008) indicates an
arbitrary relationship between Ren Qing and Relationship Harmony of the SAPI indicative of
the concept of Ubuntu outlined in Factor 1 of this solution. This scale, however, could be
revised for the South African context in order to achieve greater representativeness of one of
the two Interpersonal Relatedness factors proposed by this solution. Further research is
therefore required.
The Thrift vs Extravagancescale did not load onto either of the factors in this solution, thus
indicating that Thriftvs Extravagance is problematic in the South African context. This may
suggest that the Thrift vs Extravagance scale is not adding anything meaningful to the
collective dimension of personality in the South African context. This is supported by
previous research by Laher (2010) that found Thrift vs Extravagance to not load on any
factors for the CPAI-2.Similarly poor reliabilities were found for this scale in the current
study, as well as Cheung et al. (2008) and Laher (2010) (refer to Table 4.1).However, in the
previous six factor solution the Thrift vs Extravagance scale explained 14.46% of the
variance and loaded onto Factor 3 at .98. In this solution, Thrift vs Extravagance did not load
onto either factor, but its highest loading of .38 was on Factor 1. Hence, as discussed
previously,further research into the Thrift vs Extravagance scale in the South African context
is warranted in order to determine if a problem exists at the level of scale construction or if
this factor does not assess a collective dimension of personality in a South African sample.
The total variance explained by the two factor solution is 58.78%; with Factor 1 explaining
36.65% of the variance and Factor 2 explaining 22.14% of the variance. This therefore
suggests that Factor 1- the ability to maintain meaningful relationships- may be an important
determinant of the collective dimension of personality in South Africa and may be adding
145
more to an understanding of this dimension than Factor 2- an adherence to cultural norms and
values. This is in line with conceptions of personality in collectivist cultures that dictate that
individualsevaluate themselves in terms of others, their interpersonal roles and the quality of
their relationships (Green et al., 2005).This solution therefore provides evidence of an
Interpersonal Relatedness factor in this sample, while also indicating the value of this etic
measure in an understanding of the collective dimension of personality in South Africa;
however more research is required in order for this instrument to obtain sufficient validity in
a South African sample.
4.9. Construct Validity: SAPI and CPAI-2
4.9.1.Subscales
A joint factor analysis of the SAPI and CPAI-2 collective dimensions was examined.The
original factor structure for the SAPI collective dimension sub-clusters and the CPAI-2
Interpersonal Relatedness scales postulates twenty total subscales; however empirical
evidence suggested the existence of four factors (see Table 3.21). Hence four factors were
examined.
The SAPI sub-clusters Active Support, Encouraging Others, Guidance, Empathy, Harmony
Maintenance, ‘I help others to succeed’ and Integrity, as well as the CPAI-2 scale Social
Sensitivity loaded onto Factor 1. Factor 2 included the SAPI sub-clusters Hostility,
Approachability, Harmony Breach, Egoism and the CPAI-2 scale Harmony. Factor 3
incorporated the SAPI sub-clusters Social Desirability, Self-deception and Impression
Management. Lastly, Factor 4 included the CPAI-2 scales Traditionalism vs Modernity,
Discipline and Ren Qing.
Factor 1 of this solution is almost identical to Factor 1 of the three factor solution at scale
level for the SAPI sub-clusters. This factor, therefore, represents a collective dimension of
personality reflective of an element of Ubuntu that dictates caring for others and the
maintenance of healthy relationships, as indicative of non-Western expressions of personality
146
(Nel, 2008;Markus & Kitayama, 1998).The degree to which one is sensitive to others through
empathy, understanding, willingness to listen and communicate, and approachability- largely
encompasses the facets and definitions of the SAPI sub-clusters within this factor (Cheung,
2006; Cheung et al., 2001; Cheung et al., 2008).Nel (2008) further suggests theoretically
thatmany facets of the Relationship Harmony cluster of the SAPI that indicate socially
responsive and sensitive behaviour characterised by the Social Sensitivity scale of
Interpersonal Relatedness.For this reason the Social Sensitivity scale of the CPAI-2, as
discussed in the two factor theory suggested in the six factor solution in section 4.8.1.1.,
reflects a collective expression of the collective dimension of personality in South Africa and
thus justifiably forms part of this factor. This may therefore suggest that these elements and
qualities inherent in caring for others and the maintaining of supportive relationships of this
collective dimension of personality do in fact operate in similar ways across Chinese and
South African samples. This factor may therefore provide further evidence that the concept of
Ubuntu is more universal than expected, extending beyond the borders of South Africa to
encompass a more collectivistic element.
Factor 2 of this solution is found to be identical to Factor 2 of all the scale level factor
solutions for the SAPI discussed above. Consequently this factor can likewise be said to
reflect an opposite continuum to the collectivistic expression of personality and spirit of
Ubuntu or rather the aspects of oneself that dictate a neglect of others and a relationship
between them (Nel, 2008).Furthermore emphasising individualistic personal aims, inhibiting
the maintenance of harmonious relationships (Green et al., 2005; Markus & Kitayama, 1998).
Harmony loads negatively onto this factor thus describing an individual who possesses the
qualities of being confrontational, insensitive and open to conflict(Cheung, 2006; Cheung et
al., 1996; Cheung et al., 2001). The Harmony scale therefore loads in congruence with the
theme of the factor; assessing those qualities of individuals that inhibita care toward others
and relationship with them. This therefore corresponds to the two factor theory suggested in
the CPAI-2 six factor solution in section 4.8.1.1. above. Factor 2 can therefore be described
to possess an individual expression of the collective dimension of personality in that it
appears to assess one’s relationship with others and how the individual approaches this
relationship with their respective qualities and abilities. Thus further providing evidence that
South Africans may function in collectivistic and/or individualistic environments and thus
147
they adopt the necessary qualities of each environment depending on the situation (Nel,
2008).
Factor 3 is representative of the Social Desirability cluster and its sub-clusters of the SAPI
(similar to the three factor solution at scale level for the SAPI). Factor 3 is theorised to
measure one’s need for approval by the MCSDS, and one’s tendency to give honest but
positively bias responses; as well as one’s habitual presentation of a favourable public image
respectively by the BIDR (Leite & Beretvas, 2005; Paulhus, 2002; Uziel, 2010). The subclusters of Social Desirability have however proven to be problematic as discussed
previously, particularly the MCSDS. The inclusion of the Social Desirability cluster was done
as a means with which to assess social desirability bias; yetevidence of the overlap between
the Social Desirability sub-cluster of the MCSDS and Integrity as provided in previous
solutions indicates that Social Desirability may in fact be measuring an element of
personality distinct to other dimensions (Weiner & Craighead, 2010). Furthermore, research
on the MCSDS has indicated that Social Desirability is regarded as a personality trait more
than a response bias (Verardi et al., 2010). Thus Social Desirability may be included as an
element of the collective dimension of personality in South Africa; however further research
is required.
Lastly Factor 4 of this solutionis represented by theTraditionalism vs Modernity, Discipline
and Ren Qing scales of the CPAI-2. This factor is identical to Factor 2 of the two factor
solution of the CPAI-2 in section 4.8.1.2. above, thus providing evidence of this two factor
theory found in this study, as opposed to the two factor theory suggested in section
4.8.1.1.This factor can therefore be said to assess an element of Chinese culture or tradition in
relation to one’s adherence of cultural norms, beliefs and traditions (Cheung, 2006; Cheung
et al., 2001; Cheung et al., 2008).Confucianism forms the foundation of Chinese culture and
tradition and provides the basis for interpersonal behaviour through behavioural and moral
doctrines regarding human relationships, social structures, virtuous behaviour and work
ethics (Pye,1972; as cited in Fan, 2000). In Confucianism, rigid rules dictate human
interaction and social behaviour such that discipline is regarded very highly in Chinese
cultural tradition (Fan, 2000). These rules are distilled in the Five Constant Virtues;namely
humanity, righteousness, propriety, wisdom and faithfulness (Ch'en, 1986; as cited in Fan,
148
2000).However, this Chinese cultural tradition appears to operate in a similar way in this
South African sample, revealing its possible universality as a collective dimension of
personality.
TheThrift vs Extravagance of the CPAI-2did not load onto any of the 4 factors in this
solution, once again alluding to its potential problematic inclusion as a collective dimension
of personality in the South African context. This is further emphasised by the lack of factor
loadings for the Thrift vs Extravagance scale in a South African sample in a study by Laher
(2010); as well as the low reliabilities found for this scale in this study (.36) and Laher’s
study (.37). Research by Cheung et al. (2008) also indicated that this scale likewise possessed
a low reliability of .57; thus suggesting that this factor is not universal across cultures and
may in fact not meaningfully assess a collective dimension of personality.However, the Thrift
vs Extravagance scale loaded by itself in the CPAI-2 six factor solution, thus indicating that
further research of this scale is necessary. The Egoism sub-clusterof the SAPI was proposed
by Nel (2008) to contain facets which encompasses Thrifty behaviour characterised by the
Thrift vs Extravagancescale; however Thrift vs Extravagance did not load with the Egoism
sub-cluster in this solution. Further research is therefore required for the Thrift vs
Extravagance scale in the South African context.
The total variance explained by the factors in this model was 64.07%, with Factor 1
explaining 34.75% of the variance, Factor 2 explaining 15.62%, Factor 3- 7.03% and Factor
4- 6.67%. It is therefore evident that Factor 1 and Factor 2 may be adding a greater
understandingto the collective dimension of personality in South Africa, particularly Factor 1.
The same result was found for all three solutions at scale level for the SAPI, thus further
supporting this conclusion.
4.9.2.Scales
An exploratory factor analysis with varimax rotation was conducted on the collective
dimension scales of the SAPI and CPAI-2.The original factor structure for the SAPI
149
collective dimension scales and the CPAI-2 Interpersonal Relatedness scale postulates six
overall factors (five SAPI clusters and one CPAI-2 factor); empirical evidence, however,
suggested the existence of only two factors (see Table 3.24).
Factor 1 incorporated the clusters Social Desirability, Soft-Heartedness, Integrity and
Relationship Harmony of the SAPI.In the SAPI, the clusters of Soft-heartedness, Integrity
and Relationship Harmony are said to correspond the strongest with the description and
meaning of Ubuntu (Nel, 2008). Therefore it may be plausible to suggest that Factor 1
assesses an element of the concept of Ubuntu inherent in collectivism, in which one feels a
sense of duty to one’s group and a desire for social harmony (Green et al., 2005). In
accordance with the five factor solution of the SAPI, Social Desirability loaded with Integrity
in this solution; further emphasising that these clusters may be tapping into similar constructs.
Furthermore this indicates that aspects of the Social Desirability cluster may in fact be
measuring something akin to Integrity, as opposed to social desirability; suggesting that
Social Desirability may be more of a personality trait more than a response bias (Verardi et
al., 2010).This factor therefore reveals a strong inclination to be sensitive towards others and
to build and maintain a community; thus revealing a collective dimension (Nel, 2008).
Factor 2 included the SAPI Facilitating cluster and the CPAI-2 Interpersonal Relatedness
factor. This loading is an interesting one and one that is not easily discernible. However it
may be said that Interpersonal Relatedness captures the collective qualities of humanity and
social decorum (Cheung et al., 2008). Facilitating- on the other hand- describes an individual
that displays such qualities by guiding and encouraging others to obtain these exact abilitieshumanity and social decorum. In the cluster these qualities manifest in the improvement of
self and the overcoming of life’s challenges (Nel, 2008). For this reason, it may be argued
that Facilitating and Interpersonal Relatedness encompass elements of the same construct that
possess both individual and collective expressions of social-relatedness.This is particularly
relevant with regards to Harmony and Social Sensitivity, which contain elements of
Facilitating in their possession of inner harmony and sensitivity towards others that
encourage such facilitation. This factor emerges in line with collectivist cultures that
emphasise one’s interpersonal roles and behaviour in line with group norms in order to
maintain relationships with others (Green et al., 2005; Markus & Kitayama, 1998).
150
The combined variance for this model is 69.68%; with Factor 1 explaining 50.71% of the
variance and Factor 2 explaining 18.97% of the variance- thus indicating that Factor 1 may be
contributing more to the collective dimension of personality in South Africa. This is
particularly interesting in light of the explanations and descriptions of these two factors
provided in the discussion above, as it would consequently appear that the Ubuntu factor is
providing meaningful information into the understanding of the collective dimension of
personality in South Africa.
4.9.3. Six Factor Solution
An exploratory factor analysis with varimax rotation was conducted with the extraction of six
factors in order to determine if the original factor structure is replicated in a South African
sample. Results from Table 3.25revealed the existence of six factors; Interpersonal
Relatedness of the CPAI-2 loaded onto Factor 1, Social Desirability of the SAPI loaded onto
Factor 2, Integrity onto Factor 3, Facilitating onto Factor 4, Relationship Harmony onto
Factor 5 and finally Soft-Heartedness of the SAPI loaded onto Factor 6.
This solution provides evidence for a collective dimension of personality in South Africa, as
indicated by the subscales of the SAPI and CPAI-2.The results of the factor analyses
conducted howeverindicate that the SAPI and CPAI-2 are inappropriate for use in South
Africa as they stand and further work is required on both of these instruments in this context.
Although a collective dimension of personality is discernible; the individual and joint factor
analyses reveal that they lack construct validity. Further research is warranted in order to
determine whether congruence exists between the collective dimension factor solutions in the
South African context, as well as the order and importance of factor loadings.This study
therefore reveals that an emic measure- the SAPI- and an etic measure- the CPAI-2- may not
be applicable in the South African context in the manner in which they stand.
151
The total variance explained by this model was 100%; with Factor 1 explaining 50.71% of
the variance, Factor 2 explaining 18.97%, Factor 3- 12.29%, Factor 4- 8.78%, Factor 55.34% and Factor 6- 3.91%. The amount of variance explained in this solution may provide
evidence of the six factor collective dimension of personality in South Africa.Interestingly, in
this solution, the Interpersonal Relatedness factor explains the most variance with 50.71%considerably more than the remaining five factors. This seems to suggest that this factor is
providing a greater understanding into the collective dimension of personality in South
Africa; which stands in contrast to the two factor solution discussed in section 4.8.1.2. above.
It seems even more bizarre then that the Social Desirability cluster explains the second most
variance in this solution. As discussed previously, the Social Desirability sub-cluster was
found to load with Integrity in the five factor solution thus indicating that this sub-cluster of
the MCSDS may measure an element of Integrity as opposed to Social Desirability.
Furthermore, this MCSDS has been found in research to reflect more of a personality trait
than a response bias (Verardi et al., 2010). What seems to be most fascinating is that
Relationship Harmony and Soft-Heartedness- who are theorised to represent the concept of
Ubuntu- explain the least variance. The order of the factor loadings therefore may provide
some interesting understandings into the collective dimension of personality in the South
African context; yet further research is required in order to determine this.
South Africa is considered to be both an individualistic and collectivistic culture. South
Africans consequently may function in collectivistic and/or individualistic environments and
thus adopt the necessary qualities of each environment depending on the situation (Nel,
2008). Nonetheless, all South Africans have been shown to possess a strong inclination to be
sensitive towards others and to build and maintain a community; thus revealing a collective
dimension (Nel, 2008). As a result, the sub-groupings within South Africa were explored
further in terms of bias.The great diversity present in South Africa indicates that it may not be
homogenous or and equivalent across cultures and language groups (Abrahams & Mauer,
1999; Foxcroft, 1997; Sehlapelo & TerreBlanche, 1996; Van Eeden & Prinsloo, 1997). As a
result, bias was explored in terms of gender, race and home language for the SAPI and CPAI2.
152
4.10. Construct Bias: SAPI and CPAI-2
Construct bias was examined in this study for the variables gender, race and home language
using ANOVAs.
4.10.1. Construct Bias across Gender
ANOVA results examining gender differences across the SAPI clusters and sub-clusters are
presented in Table 3.26. From these results it is evident that,in terms of gender, males
scoredhigher in Egoism, Hostility, Approachability and Harmony Breach. Therefore, this
suggests that males are more likely to be greedy, jealous and selfish; while strongly inclined
to be physically, psychologically and emotionally aggressive as compared to females. Males
also place themselves above others, thus making them less accessible for others on an
interpersonal level;as well as causing them to form ruptures in their relationships (Nel, 2008).
Females, on the other hand, were found to score higher on Empathy, Active Support and
Harmony Maintenance. Therefore, suggesting that females are also more likely than males to
be compassionate, considerate and caring, along with generous and actively involved in the
well-being of the community. The Harmony Maintenancesub-cluster, however, provided
evidence of a significant difference with a small effect size.
ANOVA results examining gender differences across the CPAI-2 factor and scales are
presented in Table 3.26. From these results it is evident that, in terms of gender, females
scored higher on Interpersonal Relatedness factor and scale of the CPAI-2; namely Ren Qing,
Social Sensitivity, Discipline and Harmony.This indicates that females are more likely than
males to emphasise the importance of social values embedded in collectivist culture and are
more likely to adhere to cultural norms.Females are also considered to a lesser extent, slightly
more likely than males to possess inner harmony and peace of mind, in addition to being
more sensitive to others. Interestingly, females were also found to be more likely to be
disciplined and rigid than males (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001).
153
The ANOVA results thus indicate the existence of differences between males and females in
terms of certain scales and subscales of the SAPI and CPAI-2. These differences are
significant, particularly for Interpersonal Relatedness, Ren Qing, Discipline, Egoism,
Hostility, Empathy, Approachability and Harmony Breach, which demonstrate effect sizes of
.35 and above (Hyde, 2005). This therefore suggests that a gender bias may in fact exist.
However, an argument exists in which such gender differences may be considered to be a
result of fundamental differences between males and females rather than bias (Friedman &
Schustack, 2006; Larsen & Buss, 2008).Laher (2010) found significant differences for males
and females in a South African sample for Social Sensitivity, Harmony, Interpersonal
Relatedness and Thrift vs Extravagance, such that females scored higher than males on these
scales.
Despite this argument, there is still value in the bias argument as such difference needs to be
addressed and controlled for. This is particularly relevant in organisational settings were
males and females apply for the same jobs and are required to complete the same
psychometric instruments.
It is also believed that differences exist across gender in individualist and collectivist cultures
(Costa et al., 2001). Individualistic characteristics tend to be associated with masculinity;
while collectivistic traits are associated with femininity (Kashima et al., 1995; as cited in
Green et al., 2005). This analysis appears to provide evidence of this in that Males were
found to score higher on the traits that have been described to possess an individualistic
orientation such as Hostility and Egoism; while females were found to score higher on Active
Support and Social Sensitivity. However in South Africa such distinctions across cultures are
usually difficult to make. Hence race and home language were explored.
4.10.2. Construct Bias across Race
ANOVA results examining racial differences across the SAPI clusters and sub-clusters are
presented in Table 3.27. From these results it is evident that non-Whites scored higher than
154
Whites on Facilitating, Guidance, Encouraging Others, ‘I help others to succeed’, Harmony
Maintenance and Impression Management.The Harmony Maintenance and I help others to
succeed sub-clusters, however, provided evidence of a significant difference with a small
effect size.The resultsindicate, nonetheless, thatnon-Whites are more likely than Whites to
guide others actively or inactively through life’s challenges by leading and advising them
throughteaching about right and wrong, and through personal example; and to a lesser extent,
to encourage and uplift others to improve themselves (Nel, 2008).
ANOVA results examining racial differences across the CPAI-2 factor and scales are
presented in Table 3.27. From these results it is evident that higher scores were found for
non-Whites for Interpersonal Relatedness and Traditionalism vs Modernity; thus providing
some evidence for the existence of this collective dimension in the South African context.
This thus indicates thatnon-Whites are more likely than Whites to stress the importance of
one’s social values and also tend to be more conservative, obedient and protective of old
customs (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001).
When considering the many lines of segregation created by Apartheid in South Africa,
particularly for non-Whites, who this political dispensation did not serve, it would seem
almost inevitable that post-Apartheid would result in a unification or alliance of non-Whites
in order to protect themselves and their customs. This unity would in turnresult in an
emphasis on a collective collaboration- a community in which one individual helped and
supported the other. This unity therefore describes an element of Ubuntu inherent in the
collectivism dimension (Nel, 2008). Furthermore it seems sufficient to suggest that
Facilitating, as an element of collectivism and therefore Ubuntu (discussed previously), may
prove to be most rife for non-Whites as a result of a history of Apartheid or the nature of
acollectivistic culture. In addition, the resulting injustices of Apartheid could be argued to
cause those persecuted to hold onto their traditions, customs and beliefs in an attempt to
protect themselves and reclaim their identity. As discussed above, this would also result in
the unity of the persecuted people, such that a care for others, respect for one’s relationships
and an emphasis on one’s social values emerges (Laher, 2010; Meiring, 2006; Nel, 2008).
155
It is also suggested that non-Whites tend to consciously practice deception or inflate positive
results (Paulhus, 2002). Psychometric assessments, such as personality tests, were used as
tools by the Apartheid government with which to perpetuate inequalities and serve the White
elite, thus denying Black labour access to educational and economic opportunities. Hence,
psychometric tests were and still are perceived by non-Whites to be a source of bias and
unfairness (Foxcroft & Roodt, 2005; Meiring, 2006; Laher, 2010). As a result, it would seem
plausible that non-Whites attempt to redress such issues by attempting to create a positive
image of themselves through inflation of results and deception. This further emphasises the
importance of creating and utilising tests that are fair and unbiased in the South African
context (Sehlapelo & Terre Blanche, 1996; Laher, 2010; Nel, 2008). Furthermore, it has been
found that culture may influence the need for social approval, such that socially desirable
responses may differ across countries (Crowne & Marlowe, 1964; as cited in Verardi et al.,
2010).In addition, cultural differences occur around self-disclosure and the pressure to
conform to social norms and values(Verardi et al., 2010). Considering the IndividualismCollectivism dimension, it would appear plausible that collectivist cultures are more likely to
respond in socially desirable ways due to the expectation to be loyal to the group. A study
conducted by Lalwani, Shavitt and Johnson (2006) found self-deception to be associated with
individualism; while impression management was associated with collectivism.
The results of this analysis also revealed that, in contrast, non-Whites scored lower than
Whites onEgoism, Hostility, Integrity, Approachability and Harmony Breach.The Hostility
sub-clusterprovided evidence of a significant difference with a small effect size. Further
results indicate that Whites are slightly more likely than non-Whites to be greedy, jealous and
selfish; as well as place themselves above others, making them less accessible for others on
an interpersonal level and so causing fissures in their relationships. Interestingly, however,
Whites are also indicated to be more likely to possess moral soundness, honesty and
trustworthiness (Nel, 2008).As discussed above, non-Whites may be considered to be less
trusting and honest as Whites due to their history of persecution by others as a result of
Apartheid or the nature of a collectivistic culture in which members of the community are
involved in each other’s lives. The concept of Ubuntu or spirit of collective collaboration
may also be found to be lessrife with non-Whites as the emphasis on one’s relationships with
others is not deemed to be of the utmost importance, but rather one’s selfish and individual
aspirations within limit. Individualist and collectivist attitudes cantherefore be activated as a
156
function of one’s social context and social relations, such as belonging to a culture (Green et
al., 2005; Singelis et al., 1995). This may indicate an interesting finding in terms of the
individualism/collectivism distinction in South Africa’s multicultural society, such that
Whites may show more individualistic tendencies than non-Whites.
However, differences in personality between Whites and non-Whites are also argued to be a
cultural phenomenon, in which differences occur as a result of how these personality traits
are expressed in individualist and collectivist cultures (Allik & McCrae, 2004);this thus
introduces the argument of acculturation. In a multicultural society such as South Africa, in
which different cultures have come into closer contact with each other, it is argued that such
rich cultures are bound to influence one another.Acculturation therefore delineates a
modification of original cultural patterns of a particular group as a result of the continuous
direct contact experienced with this group and other cultures (Heuchert et al., 2000). This
modification is of particular relevance in a student sample in which individuals from different
cultures with an individualistic and/or collectivistic orientation come together to receive an
education in order to achieve individualistic goals such as the achievement of one’s potential
and competition (Oyserman, Coon & Kemmelmeier, 2002; as cited in Laher, 2010; Van Dyk
& De Kock, 2004; as cited in Laher, 2010). It is argued, furthermore, that acculturation could
in fact go as far as to influence the cultural differences that occur in a society or if they occur
at all (Eaton & Louw, 2002). South Africa has been described as experiencing an ideological
shift toward Western individualism, thus dictating an increasing individualistic orientation
over a collectivistic orientation (Mpofu, 2001). Therefore due to the use of a student sample
in this study, the acculturation argument should be considered and kept in mind.
The Thrif vs Extravagance scale was found to replicate in the six factor solution of this study
in section 4.8.1.1.; however it was not found to replicate for any other solution, as in Laher
(2010). When considering this scale and what it aims to assess, one may expect differences to
occur between Whites and non-Whites on this scale, particularly in light of the consequences
of Apartheid. The lack of educational and economic resources provided by Apartheid for the
non-White population resulted in a lack of opportunities and poverty for this race group
(Sehlapelo & Terre Blanche, 1996; Laher, 2010), such that differences in Thrift
vsExtravagance behaviour may be expected. However, the lack of evidence for this expected
157
difference in this study leads one to wonder if South African society has overcome this; as
higher socioeconomic classes have been found to exhibit more individualistic behaviours
(Freeman, 1997; as cited in Green et al., 2005). A conclusion is difficult to ascertain in such
an acculturated student sample as in the current study, due to the growing affluent African
population that may be argued to be represented in this sample.
4.10.3. Construct Bias across Home Language
ANOVA results examining differences across home language for the SAPI clusters and subclusters are presented in Table 3.28. From these results it is evident that English second
language speakers scored higher than English first language speakers on ‘I help others to
succeed’, Facilitating, Guidance, Encouraging Others and Impression Management.The ‘I
help others to succeed’sub-clusterprovided evidence of a significant difference with a small
effect size.Furthermore, the results indicate that English second language speakers guide, lead
and advise others on life’s challenges through teaching about right and wrong and through
personal example; while they are more likely than English first language speakers to
motivate, encourage and uplift others in their endeavors.This indicates that English second
language speakers, similar to non-Whites, transcribe to a more collectivistic orientation
dictated by collective collaboration and the spirit of Ubuntu (Nel, 2008). However, English
second language speakers are also revealed to be slightly more likely to inflate positive
results similar to non-Whites- suggested to be a result of a distrust in psychometric testing
and others due to the inequalities and bias present in the previous political dispensation
(Foxcroft & Roodt, 2005; Laher, 2010; Nel, 2008; Sehlapelo & Terre Blanche, 1996).
However these differences may also just prove to be a result of English second language
speakers taking the test in English.
English second language speakers, however, scored lower than English first language
speakers on Integrity and Approachability. The Approachability sub-clusterprovided
evidence of a significant difference with a small effect size. Thus suggesting that to a lesser
extend English first language speakers are more likely to possess moral soundness, integrity,
honesty, trustworthiness and responsibility than English second language speakers.
158
ANOVA results examining home language differences across the CPAI-2 factor and scales
are presented in Table 3.28. From these results it is evident that English second language
speakers scored higher than English first language speakers on Interpersonal Relatedness
andTraditionalism vs Modernity. These results thus suggest that English second language
speakers are slightly more likely to emphasise the importance of one’s social values than
English first language speakers, and tend to be more conservative, obedient and protective of
old customs (Cheung, 2006; Cheung et al., 1996; Cheung et al., 2001). This cling to one’s
social values, customs and traditions could be argued to be a result of the persecution of
Apartheid in which such individuals were torn from the essence of their identities; however
this could also be a result of the nature of collectivistic cultures.Nonetheless, this therefore
provides evidence of a collectivist orientation for English second language speakers of this
sample.
These results highlight an interesting similarity found between English second language
speakers and non-Whites in this study in terms of the Traditionalism vs Modernity,
Facilitating, Guidance, Impression Management, Integrity and Approachability dimensions.
While both the race and home language variables are valuable to examine, there is evidence
here to suggest that these variables may be tapping into different expressions of the same
variable. The similarities found between English second speakers and non-Whites provides
evidence of a difference in personality for collectivist cultures- particularly for the
Traditionalism vs Modernity, Facilitating and Guidance dimensions- as English second
speakers are believed to subscribe to more collectivist than individualist ideas. This also
illustrates the argument that Africans and Asians appear to have similarities in their
expression of personality (Laher, 2010).Further research is required in terms of bias in home
language in order to determine whether bias was in fact operational or whether we should be
testing for genuine differences.
4.11. Thematic Content Analysis
159
Two focus groups were conducted in this study in which seven themes emerged from the data
obtained. These themes include a general understanding of personality, the innate existence
of personality vs the over expression of personality, collectivism in personality, a national
identity, culture, psychometric testing and social desirability. These themes will be discussed
below in detail in the sections to follow.
4.11.1.General Understandings of Personality
Personality according to participant F2 is said to be“the portrayal of peoples’ individualities”,
such that its definition cannot be understood in terms of labels or categories, but that
personality is rather unique to each individual experience and inner qualities representing the
distinctiveness between individuals (Weiten, 2007).Personality can thus be defined as “the
dynamic organisation within the individual of those psychophysical systems that determine
his unique adjustments to his environment” (Allport, 1937; Robbins, Judge, Odendaal &
Roodt, 2009, p. 92).
From the early times of the theatre, in which personality was depicted as an array of different
masks, to modern conceptualisations of personality as the core self- representative of one’s
inner qualities and behaviours- personality has come to possess many diverse understandings
(Laher, 2010; Schultz & Schultz, 2008). Thrusting this diversity further, is the existence of
numerous definitions of personality; the term personality in psychology being said to have as
many definitions as there are theories (Kline, 1993). Eight theoretical approaches to
personality exist in the literature; namely psychodynamic, the lifespan theories, cognitive,
social learning, humanistic, behaviourist, biological and the trait approach;each approach
depicting a conceptualisation of personality and a means with which to assess it.
Consequently individual conceptualisations of personality result from an individual frame of
reference that ascribes human behaviour to innate aspects of the individual, such that a
number of definitions and conceptualisations can be provided to illustrate the core of
personality (Laher, 2010;Larsen & Buss, 2008; Schultz & Schultz, 2008). Some of these
conceptualisations were presented in the focus group.
160
A topic of major debate within the focus groups with regard to a general understanding of
personality was the distinction between personality and character. The common
understanding indicating that character is believed by members of focus group 2 to be “the
essence of someone” as stated by participant F3; while personality, on the other hand, is
believed to be how one views the world, how they are shaped by it and how they represent
themselves to the world as a result- thus indicating a broad understanding of personality. This
understanding of personality and character thus suggests a move towards an understanding of
traits rather than types of personalities. Traits represent the “core of personality” and are
determined by one’s habits, attitudes, skills, beliefs, and other characteristic adaptations
(McCrae, 2001; as cited in Laher, 2010). For this reason traitsare thus expressed in stable and
enduring individual differences in thoughts, feelings and behaviour (Church, 2000), that
manifest from one’s historical, cultural and social context (McCrae & Costa, 1995).
4.11.2.The Innate Existence of Personality vs the Overt Expression of
Personality
The focus groups of this study indicated personality to be a combination of innate personality
variables explicated from birth, as well as variables contracted by one’s overt environment.
Participant F2, in particular, explained personality to be something you are born with, but
which is moulded by your experiences and the people around you, such that “some people are
just a product of where they come from [overt], whereas other people are just born that way
[innate]”.Personality as defined by trait psychologydescribes consistent patterns of individual
feeling, cognition and behaving (Church, 2000; John, Robins & Pervin, 2008) resulting from
an innate existence of personality. Personality is considered to be an organisation of traits that
interact, rather than the existence of traits in isolation (Freeman, 1962), such that these innate
individual traits exhibit predispositions to certain behaviours(Tellegen, 1991; as cited in
McCrae & Cost, 2008b). Thus the overt expression of these innate personality structuresstem
from one’s historical, cultural and social context (McCrae &Costa, 1995).
This paradox within the origins and expressions of personality thus indicates that individuals
possess predispositions in terms of the way that they behave, as a result of their unique traits
that define who they are; yet these predispositions manifest as a result of one’s environmental
161
circumstances (Church, 2000).Traits can therefore be declared to be situation-dependent
(Tellegen, 1991; as cited in McCrae & Cost, 2008b), in that a “person’s personality changes”
as proposed by Participant J2. It is thus suggested that the overt expression of one’s
personality is subject to change as a result of environmental circumstances; in that individuals
adapt the personality traits that they overtly express to others in specific situations, however
one’s innate personality remains fundamentally intact (Participant F4). Five Factor Theory
(FFT), similarly, proposes that one’s Basic Tendencies and Characteristic Adaptations are
central to personality. Basic Tendencies represent the psychological potentials of the
individual that are rooted in biology and so are not susceptible to changes from the
environment (Allik & McCrae, 2004; McCrae & Costa 1996) - thus describing the innate
existence of one’s personality. Characteristic Adaptations are, conversely, acquired through
the individual’s interaction with the environment and are consequently susceptible to change
through external influences (Allik & McCrae, 2004; McCrae & Costa 1996). This may be
likened to the overt expression of one’s personality, as mentioned above. The Self-Concept is
a Characteristic Adaptation shaped by one’s Basic Tendencies that describes a view of
oneself as a result of life experiences and social feedback. This aspect of personality seems to
provide a combination of the innate existence and overt expression of personality, and is of
particular importance in personality assessment (Allik & McCrae, 2004).
In addition to being adaptable, personality is indicated by participants in the focus group to be
learnt from one’s background, upbringing and history (Participant J2). Personality is believed
to transpire from one’s historical, cultural and social context (McCrae & Costa, 1995). This
reveals that one “cannot ignore the influence of past experiences, as well as environmental
factors; like where you grew up or the cultural factors that affect your personality,” as
indicated by participant J3. As a consequence of the influence of one’s environmental
contexts and experiences, one’s social context is emphasised to be of importance- revealing
that one’s community (and “the concept of the spirit of the community”- participant J2) is
instrumental in moulding individual personality. Thus, this provides an argument for the
existence of Ubuntu and a collective dimension of personality. Furthermore, this argument
emphasises the culturally-conditioned phenomena of the FFT incorporated inCharacteristic
Adaptations and External Influences within the personality system(Allik & McCrae, 2004;
McCrae & Costa, 1996; 2008b); thus indicating that it is here that the collectivistic dimension
is submerged. In addition, this may further suggest that one’s Characteristic Adaptations and
162
External Influences dictated by one’s culture and its subsequent norms (Allik & McCrae,
2004; McCrae & Costa, 1996; 2008b) are more important and influential in determining
one’s personality in the South African context than postulated by FFT.
4.11.3.Collectivism in Personality
The individualistic collectivistic dimension determines whether the set of traits that
individuals within a particular culture use to describe themselves centre around
individualistic or societal concerns (Green et al., 2005). It has been suggested that western
cultures are representative of individualism with an egocentric focus on “I” and “me”; while
non-western cultures are indicative of collectivism with a sociocentric focus on “us” and
“we” (Triandis, 1995; Green et al., 2005). Personality in South Africa was described by the
focus groups to be a fusion of collectivistic and individualistic traits, resulting from western
and non-western influences. Thus western cultures are believed to maintain their
independence through expressions of their individuality and inner attributes rather than group
norms; while non-Western cultures, seek to maintain harmonious relationships assimilating in
to the collective group (Triandis, 1989; Markus & Kitayama, 1998). Therefore, the
dispensation of one’s culturedetermines whether the individual or society is of utmost
importance; consequently “influencing who we are and what we do” (Participant J2).These
findings correlate with those found in a South African sample, in which the English language
group tended to describe others in conceptualised terms such as patient, caring or respectful
(individualistic), whereas the Siswati language group used more socially descriptive terms
like, she is patient with her husband or she always greets elderly people (collectivistic) (Nel,
2008).Personality assessment in the South African context will assist in the clarification of
this issue.
The collective construction of personality refers to relationality, while individualistic
constructions view and describe groups in terms of individual identities submerged within a
social identity (Ellemers, Spears & Doosje, 2002; Laher, 2010; Markus & Kitayama,
1998).The focus groups of this study indicated the existence of a collective dimension of
personality in South African, due to the focus on family and community, and the progression
of a communal united nation.Collectivist personalities are therefore determined by
163
interpersonal situations (Cross & Markus, 1999; as cited in McCrae, 2004).At the same time,
however, a fusion between collectivism and individualism in personality exists as South
Africa is experiencing a move towards westernisation in which “society dictates that we must
be an individual” (Participant J4).The focus group additionally suggested that it is within
one’s cultural group that one’s individual traits emerge;such that South Africans may
function in collectivistic and/or individualistic environments and may adopt the necessary
qualities of each environment depending on the situation (Nel, 2008).Thus according
toparticipant F2 of the focus group “[these] elements… depend on where you are and also the
community that you are in.” This may be said to be a result of acculturation- the modification
of original cultural patterns of different groups as a result of their continuous direct contact
with other cultures (Heuchert et al., 2000); such that the existence of cultural differences is
brought into question or rather merely the degree to which these differences are expressed
(Eaton & Louw, 2002). Berry and Sam (1997) propose a bidimensional model of
acculturation that depicts four types of acculturation; namely integration, separation,
assimilation and marginalisation. These four types of acculturation relate to whether the
individual wants to maintain a good relationship with the dominant or host culture
(adaptation), or whether the individual wants to maintain a good relationship with their native
culture (Berry & Sam, 1997).
Different cultures have been described as possessing individualistic and/or collectivistic
traits; thus determining whether the individual or society is of utmost importance (Green, at
al., 2005). However, it is within one’s community that this is permitted to transpire.
Consequently the type and location of one’s community- rural or urban- was identified by the
focus groups as playing an important role in determining the degree of collectivism and
individualism of the culture that shapes personality. Participants emphasised this point by
providing examples such as “a Pedi in Polokwane [is] different from being a Pedi in Jo’burg”
(Participant J8). One’s cultural identity is therefore suggested to be influence by one’s urban
or rural community; thus emphasising one’s cultural, historical and social context in
determining personality (McCrae & Costa, 1995). The focus groups suggested that the degree
to which one adopts their cultural lifestyle is a result of the type of community and the related
expectations emergent from that community; such that these expectations are more
pronounced for one’s culture in rural areas. Consequently, participant F5, argued that it is the
place where you reside that dictates whether you subscribe to individualistic or collectivistic
164
idealsand not entirely your cultural group. It is believed that modern industrial-urban cultures
tend to me more individualistic; while traditional agricultural-rural cultures tend to be
collectivistic (Singelis et al., 1995).
4.11.4.A National Identity
National identity is defined as “personality traits that are perceived to be prototypical of
members of a culture” (McCrae et al., 2005b, p. 408). It is the array of personality traits that
individuals within a culture use to describe the national character of themselves and their
neighbours (McCrae, 2002; as cited in Laher, 2010). It is this national character that reflects
emotional, attitudinal, interpersonal and motivational styles of that particular culturerevealing a distinct national identity (Terracciano et al., 2005b; as cited in Laher, 2010). The
term ‘Ubuntu’ has been used to describe a ‘South African’ national identity, thus
encompassing an interpersonal element and collective spirit of South Africans (Nel,
2008).The focus group indicated that one’s definition of the term ubuntu differs across
cultures and individuals. Participant J6 suggested this concept to mean “Community”, such
that one’s conceptualisation of community and expressions of ubuntu are different across
cultures. Participant J4 argued that in South Africa we need to extend the boundaries of
ubuntu to surpass one’s community and encompass the whole nation, such that the concept of
ubuntu in South Africa taps in to the pride and spirit reflective of one’s nationality.
All South Africans have been shown to possess a strong inclination to be sensitive towards
others and to build and maintain a community; thus revealing a collective dimension (Nel,
2008).Ubuntu is an indigenous concept that originated in South Africa to describe a ‘South
African’ way of thinking (Nel, 2008).The term Ubuntu (humaneness) possesses a
collectivistic orientation and describes a person as a person through others (Louw, 2002; Nel,
2008).However, some debate exists, in literature and within the focus groups, as to whether
ubuntu is a concept unique to South Africans. It is largely suggested that this concept of the
community indicative of ubuntu is in existence in other nations around the world, such as
“Mediterranean cultures” (Participant J6); as well as in Chinese culture in which
Interpersonal Relatedness scales were found to replicate well in a South African sample in
this study. Ubuntu is thus said to be common in all language groups in South Africa, though
165
not exclusive to them (Nel, 2008). Ubuntu as described by Louw (2001; Laher, 2010) is not
conclusively a collectivistic dimension of personality, but rather it describes the functioning
of an individual within a community, thus emphasising the community.
Focus group members highlighted the importance of our history and our past in shaping our
future in South Africa.The political and economic dispensation of South Africa is believed to
influence its personality. Apartheid presented personality in South Africa as distinct
personality traits for different cultural groups; however at present a communal national
identity is developing from an amalgamation of cultures. It is in South Africa’s multicultural
society that we are attempting to breakdown the barriers that separate us in order to develop
our own identity- one that incorporates western and non-western elements. It is this,
according to participant F2, that sets us apart from the rest of the world.The South African
lifestyle- its morals and beliefs- differentiate us and shape who we are, such that South
Africans possess a friendliness, humour and ability to laugh at themselves that is replicated in
no other nation creating language and cultural barriers with the rest of the world. Although
the South African personality is unique in its inclusion of and adaptation to multiple cultures,
we have also begun to adopt aspects of westernisation as a result of growth and the media.
Participant J3 argued that “Urbanisation brings forth westernisation and westernisation
doesn’t embrace ubuntu.”
4.11.5.Culture
South Africa is a “rainbow” nation inclusive of an array of races, languages and cultures
combined to represent one nation (Franklin, 2009). Every culture encompasses a set of norms
and values in which individuals from that culture aspire to adhere to; it is these cultural
specificities that shape personality thus finding expression through universal structures of
personality (McCrae, 2002a; as cited in Franklin, 2009; McCrae et al., 2005).Culture and
personality are thus seen to be inseparable elements (Rolland, 2002; as cited in Franklin,
2009).The focus groups of this study indicated a difference in personality between the
different cultures within South Africa, such that one’s culture influences their personality.
Participant J7 arguedthat one’s culture dictates how one is supposed to behave within that
culture and it is this that influences personality. Therefore culture is a very influential aspect
166
in this respect, as it determines how we respond to certain situations (Larsen & Buss, 2008;
Schultz & Schultz, 2008).
On the other hand, participants believed that no one person has the same personality, but
instead possess traits and characteristics that are in common with certain groups.
Accordingly, individuals who possess these similar traits tend to stick together, such that if
you are Zulu and are surround by Zulu family and friends these ‘Zulu’ traits emerge,
influencing and moulding who you are (Participant F5).This indicates a collective experience
of personality resulting from one’s cultural context (Nel, 2008).It was further argued that the
personalities of black cultures are similar in some respects as they possess similar attitudes,
opinions and mind sets- for example in terms of the way they treat women as inferior to men
(Participant J5). Thus, personality is believed to not be differentially influenced by one’s
culture, but rather one’s beliefs and ideas which in turn influence personality.This is
especially true in South Africa where strong collective experiences result in a ‘South African
personality’ (Franklin, 2009). This argument therefore gives merit to FFT in that personality
is said to form from an interaction between an individual’s Basic Tendencies or biological
traits and the individual’s Characteristic Adaptations or environmental influences (Allik &
McCrae, 2004; McCrae & Costa, 2008b).
However, this collective personality perpetuates the development of stereotypes and
generalisations, in which people are placed into classifications in which they identify with
certain ideas or traits in order to better understand these cultures or groups.Values and beliefs
are entrenched in one’s cultural group and may lead to shared perceptions about
characteristics of culture members (McCrae et al., 2007). However, McCrae, Terracciano,
Realo & Allik (2007) provide empirical literature that discredits the stereotyping of a national
identity by arguing that judgements of national character structure (as a result of personal or
second-hand experience)do not reflect actual differences in personality traits. In fact, research
has found that there does not appear to be any truth in most national stereotypes (Terracciano
et al., 2005; as cited in McCrae et al., 2007). This inaccuracy may be ascribed to limited firsthand experience with members of a particular culture or to the collective wisdom of the
majority who lack such experience (McCrae et al., 2007).Cultural groups are said to be
similar in personality with regard to their fundamentals goals; however small differences in
167
their approach set them apart (Participant F5). This argument provides evidence for the FFT
and trait approach to personality, indicating overarching traits that describe all individual
personality. Basic Tendencies, as presented in FFT, describe traits or innate biological
structures that are not susceptible to changes in the environment (Allik & McCrae, 2004;
McCrae & Costa, 1996; 2008b).It is believed that personality structure may be universal;
however it is expressed in different ways in different cultures (Meiring, 2006). This further
indicates that one’s external influences indicative of their Characteristic Adaptations such as
cultural norms and attitudes interact with these Basic Tendencies to form one’s personality
(Allik & McCrae, 2004; McCrae & Costa, 1996; 2008b).
4.11.6.Psychometric Testing
A number of common issues arisefor the SAPI and CPAI-2 with regards to psychometric
testing. The first issue explored indicated the problematic effects of the length of the SAPI
and CPAI-2 questionnaires.A variety in questions enlarges the scope of what is measured; in
turn allowing for more reliability and precision (Spector, 1992). However a large number of
items (n= 364) increases the length of the questionnaire thus diminishing the quality of
participant responses, as individuals pay little attention and may not read the instructions
carefully or might answer in a systematic way in order to complete the test. This will thus
result in a questionnaire that does not truly reflective of the individual’s personality.
Therefore participant J4 suggested shortening the length of the scale in order to minimise the
length of time required to complete the questionnaire; however one needs to consider the
accuracy and depth of the rating scale used.
The second issue explored the repetition of some of the items of the SAPI and CPAI-2
questionnaires, thus contributing to the length of the instruments. The aim of such repetition
lies in the replication of certain responses such that the results present an accurate reflection
of one’s personality. However such repetition could have resulted in systematic responses;
while the rephrasing could have elicited different responses by the same participant, as a
result of mood or misinterpretation thus undermining the quality of the results produced by
the questionnaire. This may result in random error, which is any source of variance that
reduces the validity of a scale score. Random responding or deliberately marking answers
168
without regard to the content of the item is the most obvious source of random error.
Carelessness or misunderstanding of items can also contribute to random error (Piedmont et
al., 2000). “Random responding is most likely to occur when respondents are not motivated
to complete the questionnaire accurately,” (Costa & McCrae, 1992, p. 187).
The third issue explored by the focus groups is the use of rating scales. The SAPI and CPAI-2
adopted a 5-point rating scale, in which it was suggested that the ratings used were dependent
on the context in which the items are presented. Participants indicated the possible
inconsistency in ratings across time and context, as well as individual apprehension in
providing personal information. This may result in a continuous neutral option, thus
undermining the value of the questionnaire.This could reveal a possible method bias with
regards to the rating scale and its applicability in the South African context.Participant J3
suggested the reduction in options on the rating scale from 5 to 2 in order to eliminate the
difficulties provided by neutral responses.The rating scale of the CPAI-2 was changed to a 5point rating scale in order to achieve uniformity in response scales across both the SAPI and
CPAI-2 (V.H. Valchev, personal communication, December 14, 2011). However, the focus
group suggests that the rating scale of this instrument should resume back to its original
structure- true and false.However, one needs to acknowledge the limitations inherent in a
forced choice rating scale.Those individuals with ambivalent feelings towards the item will
be forced to be included in either of the two groups which may not accurately reflect their
response or opinion or consequently oversimplify it, therefore resulting in inaccurate
precision (Spector, 1992). The benefit of the scale options of ‘agree’ and ‘strongly disagree’
are acknowledged for items with conclusive terms such as “always” or “never” (participant
J4). Suggestions were provided from focus group participants in order to improve the quality
of the questionnaire and its responses. Such suggestions included developing a more
qualitative open-ended method to allow for a more accurate and diverse description of one’s
personality (participant F1), as well as an “it depends” option in order to account for the
ambiguous and confusing items (participant F4).
Most rating scales offer between four and seven response choices; this allows participants to
indicate which of several response choices best reflects their response to an item. Simple two
option rating scales (such as yes-no or true-false) are inconsistent, unreliable and restrict
169
measurement to only two levels, thus making it difficult to distinguish amongst these levels
(Spector, 1992).Greater response options may be more effective for use in organisations
allowing for greater distinction amongst individuals on more levels, thus providing a means
with which to more effectively deal with diversity and transformation in South Africa
(Swartz and Davies, 1997; Nel, 2008).Collectivistic cultures are believed to value monitoring
one’s own and others behaviour in order to conform to valued norms, such that they have a
more accurate recall of their own behaviour and are less influenced by the wording of items
(Ji, Schwartz & Nisbett, 2000; as cited in Sanderson, 2009). Studies have revealed that
collectivistic cultures, such as China and Japan, are more likely to use the midpoint of rating
scales and avoid extreme responses than individualistic cultures, such as America and
Canada; reflecting a tendency to fit into the group and to maintain harmony rather than
express individual differences (Chen, Lee & Stevenson, 1995; as cited in Sanderson, 2009).
An increase in the number of response choices increases precision; thus it may be said that
rating scales offering five or seven response options are better for collectivistic cultures
(Sanderson, 2009); however this is contradictory to what the students of the focus groups
suggested. It is therefore possible that the student sample in this study is more Westernised
due to the influence of acculturation and thus are able to reflect individual differences in any
rating scale.It has also been suggested, however that rating scales may be less reliable and
valid for collectivistic cultures introspecting about one’s attributes and knowing one’s selfconcept are less clear than in individualist cultures. These measures may also predict
behaviour less well as behaviour is determined by social roles, relationships and norms
(Markus & Kitayama, 1998;Walford, Tucker & Viswanathan, 2010). In collectivistic
cultures, such as China, it is not considered acceptable to give yourself a positive self-rating
as opposed to individualistic cultures (Silzer, Silzer & Dowell, 2009). More research is
needed in this regard.
Culturally and linguistically problematic items were found in the SAPI and CPAI-2,
indicating that these items need to be addressed and explored further. Certain items were
found to be culturally problematic in that they made reference to specific cultures, thus
making it difficult for every individual of every culture to relate. An example of such an item
included that of ancestral sacrifices. Such items are therefore only meaningful for the culture
170
for which it was intended. Certain items were found by participants to be culturally biased to
certain races or cultures, such as “If the man does not earn the highest salary in the house, he
should still be the head of the house”. Particular items were also found to be culturally
significant in terms of their cultural and socio-economic consequences, such as
branding.Such items were said to be culturally significant as the way one interprets them is
determined by their community and degree of urbanisation; consequently suggesting that
materialism and success is expressed differently in rural and urban communities. A number
of items were identified by participants to harbour a Chinese influence, as such concepts have
no place in Western cultures and traditions. Although these problematic items need to be
addressed, they have been developed as means with which to measure indigenous aspects of
the collective dimension of personality indicative of one’s culture and language (Nel, 2008).
Participants of the focus groups identified a number of linguistic issues with regards to the
phrasing of certain items, such that words like “some” created vague and ambiguous items
allowing for misinterpretations and inconsistenciesin meaning and responses across
individuals thus limiting the use of the instrument (Spector, 1992). Similarly, specific terms
such as “usually”, “always” or “never” created difficulties in answering conclusively with
strongly agree or strongly disagree.Participant J3 also indicated the importance of context in
interpreting the language used in certain items. It was also suggested that a number of
eloquent words were used creating potential problems with regards to understanding their
meaning- such as “filial piety”.Thus, indicating that a problem may exist in terms of the
appropriateness of some of the items for the population using the instruments, such that the
intended meaning of these items may be misunderstood. This is of particular relevance in
South Africa where the majority of the population cannot speak English or understand the
language and are illiterate, as a result of the inferior education provided for the Black
population in Apartheid (Abrahams, 1996).It was also indicated that upbringing and the way
one has been taught to speak may have a grammatical interpretation and understanding as
well.Thus education level is believed to affect one’s capacity to understand difficult items.
An individual’s socio-economic status is linked to their access to higher education and
consequently to advanced English. This would then determine whether the individual will be
able to comprehensively deal with the advanced English present in some items of the
questionnaire (Franklin, 2009). In addition to this, participant’s personality traits are believed
to vary as a function of educational level (Zhang & Akande, 2002).
171
The relevance of items within the questionnaire indicated an incongruence in the intention
and importance of some of the items in terms of personality, for example “I frequently buy
snacks”.Participants also indicated the use of English expressions or colloquial terms to be
problematicin a multicultural South African sample due to the loss in meaning or possible
misinterpretation; particularly for those whose first language is not English- further limiting
the use of the instruments in terms of population and time (Spector, 1992).An example of
such an item includes the expression “blow their trumpets”.
4.11.7.Social Desirability
A major concern with regards to personality inventories and bias is the influence of social
desirability (Spector, 1992).A number of definitions may be used to describe the concept of
social desirability; these include “the tendency to provide socially desirable responses to
statements in self-description (Edwards, 1957), to give culturally sanctioned and approved
responses (Crowne & Marlowe, 1960), to describe oneself in terms judged as desirable and to
present oneself favourable (Jackson, 1984), and to give responses that make the individual
look good (Paulhus, 1991),” (as cited in Weiner & Craighead, 2010, p. 1628). All these
definitions allude to social desirability as a response style separate to that of the personality
dimension, such that a high score could mean a high score on that dimension or a tendency to
represent oneself positively (Weiner & Craighead, 2010).
It was argued by the focus groups that social desirability is affected by one’s circumstances
and experiences, in that we are socialized in terms of how to behave and what to say in
public; while one’s experiences also shape their realities and their perceptions of the world.
Thus one’s ability to engage in social desirability is dependent one’s ability to know and
judge oneself with the honesty expected. Social desirability is also dependent on individual
familiarity with psychometric test as well as level of education;it is therefore a function of an
individual’s cognitive ability and years of education, such that these individuals will thus
possess better test-taking skills and will be more ‘test smart’, which may result in more
distortions (Furnham, 1986; as cited in Ones, Reiss,& Viswesvaran,1996).A social pressure
172
was also said to exist when answering the questionnaire as it was believed to be a class
exercise in which participants who did not complete the questionnaire would be scolded or
stigmatised for doing not so.
Certain items were found to be difficult to answer with complete honesty by the focus groups
as these items were considered to be sensitive in nature or to exude possible negative
connotations; such as the SAPI Approachability item- “I am better than others”. Some items,
such as “There is no stigma about marrying a divorced person”, from the CPAI-2
Traditionalism vs Modernity scale, created negative emotions within participants in that they
were a direct stimulus for painful or sensitive events and were therefore challenging to
answer. Personally sensitive may evoke defensiveness in participants resulting in items
possibly being subject to biasing factors (Spector, 1992).
It was suggested by the focus groups that the wording of repeated items may result in
participants answering in more socially desirable ways, as one item may be worded with
more positive connotations than the other. Participants indicated that items with negative
connotations, such as “I insult people” realistically require an answer of ‘yes’; however one is
compelled to answer in a socially desirable manner because it is often not the usual way that
one conducts themselves and therefore they do not want to create an unwarranted negative
image of themselves.Participants thus responded in such a way so as to present themselves in
the best possible light. In an article by Uziel (2010) this desire was distinguished as the
defensiveness approach and the adjustment approach. The defensiveness approach dictates
that the motivation that drives socially desirable behaviour is not social approval, but rather
the avoidance of social disapproval in turn protecting a vulnerable self-esteem. Forthwith it
can be argued that individuals high in social desirability experience life as defensive,
avoidant and inhibited. The adjustment approach, on the other hand, argues that socially
desirable behaviour is associated with emotional stability, friendliness and enhanced wellbeing in social situations in which individuals are highly motivated and highly competent
(Uziel, 2010).
The concept of social desirability has inspired many debates over the years with regards to its
definition, its pervasiveness, problems with the interpretation of test results and methods that
173
it can be controlled (Weiner & Craighead, 2010).Research on social desirability also indicates
massive debate over its legitimacy, with some authors indicating that a failure to control for it
affects the validity of one’s responses; while others suggest that social desirability is in itself
a personality variable (Paulhus, 2002; Weiner & Craighead, 2010). However, if social
desirability is a feature of many personality dimensions then social desirability is not truly
conceptually independent (Weiner & Craighead, 2010). This may be the reason for the lack
of statistical evidence to support the view of social desirability scales as a reliable validity
measure or one that detects bias reporting (Piedmont et al., 2000; Uziel, 2010). Similarly this
study found poor reliabilities for the Social Desirability cluster and sub-clusters (see Table
4.1.). Furthermore, the MCSDS Social Desirability sub-cluster loaded with Integrity, thus
indicating that it may be measuring an aspect of Integrity and possibly a personality trait
rather than social desirability bias (Verardi et al., 2010).
Social desirability can affect the construct validity of an instrument, thus it is essential that
measures of personality traits be free from extraneous variables like social desirability. This
extraneous variable has a tendency to establish irrelevant variance thus decreasing construct
validity, regardless of whether it represents error or trait variance (Ellingson Sackett &
Hough,1999). Thus social desirability corrections will allow for the removal of extraneous
variance and the better interpretation of personality scores (Ellingson et al., 1999).
Despite the ongoing debate, social desirability as a form of bias is considered to be
troublesome foruse in organisations, such as in selection procedures, as these measures have
the potential to affect conclusions (Ellingson, et al., 1999; Levin & Montag, 1987; as cited in
Ones et al., 1996). “Organizations contemplating the use of personality measures in their
selection systems, in light of the validity evidence, view the potential influence of applicant
faking as a detriment that makes it difficult to include these measures in selection decisions,”
(Ellingson et al., 1999, p. 155).Response distortion includes “Frankness, social desirability,
claiming unlikely virtues, denying common faults and unpopular attitudes, exaggerating
personal strengths, good impression, self-enhancement, and faking,” (Ones, et al., 1996, p.
660). Social desirability is a response distortion that incorporates two factors that include
self-deception and impression management (Ellingson et al., 1999).In such circumstances,
such as selection procedures, applicants are ranked based on their observed scores in these
174
proceedings and organisations then select the top ranked applicants for hiring (Ellingson et
al., 1999). Individuals therefore distort their responses in a positive direction in an attempt to
artificially increase their ranking, thus placing those that answer honestly at an unfair
disadvantage (Li & Bagger, 2006). In this wayfaking has the potential to change applicant
rankings and consequently selection decisions (Ellingson et al., 1999, p. 156).
Various methods for coping with the effects of social desirability have been proposed. A
forced-choice response format could be used such that response options could then be
matched for social desirability (Weiner & Craighead, 2010; Uziel, 2010). Test instructions
could be tailored to reduce the likelihood of social desirable responding, such as indicating
the presence of a validity index that detects social desirability (Weiner & Craighead, 2010;
Uziel, 2010). Social desirability could be statistically removed from individual scores, by
adjusting initial scores based on the social desirability scale scores (Weiner & Craighead,
2010; Uziel, 2010).Finally, items could be selected for a scale based on appropriate content
representative of the construct, such that items reflecting inappropriate response bias (social
desirability) are removed.This is the method that was adopted by the developers of the SAPI
in order to control for the effects of social desirability. The low reliabilities of the Social
Desirability cluster, as well as the loadings of the MCSDS Social Desirability cluster with
Integrity indicate that this method was not successful in its aims. In addition, given the results
of this study it would appear that controlling for social desirability bias may not be a good
thing as Social Desirability may in fact be a personality trait rather than a social desirability
bias (Verardi et al., 2010). Further research is required.
4.12. Conclusion
This chapter discussed the results obtained in this study pertaining to the utility of the
collective dimension of personality in the SAPI and CPAI-2. This was done by exploring the
reliability, validity and bias of the two instruments. It is evident from the results that the
internal consistency reliability coefficients for the SAPI scales are all very good, exceeding
.60, with the exception of Social Desirability and its sub-clusters. The internal consistency
reliability coefficients for the Interpersonal Relatedness factor and scales of the CPAI-2 were
found to be comparable to the normative study, as well as Laher’s (2010) study. Interpersonal
175
Relatedness, Traditionalism vs Modernity vs Modernity, Social Sensitivity and Harmony
exceeded .60; while Ren Qing, Discipline and Thrift vs Extravagancewere below .60 and
therefore appear to be problematic.
The SAPI did not demonstrate good construct validity, with the four and five factor solutions
not replicating the factor structure postulated by Nel (2008). However, the three factor
solution was found to be the most appropriate consequently making the most sense
theoretically and empirically. Furthermore, although the four factor solution was considered a
better fit, the two factor solution was found to present a clear distinction between the
Individualism/Collectivism and Vertical/Horizontal distinction. An item level analysis
revealed that items did not tap into the constructs that they were intended to measure for the
most part, with the exception of the majority of the Facilitating items. It is evident that the
Soft-Heartedness and Relationship Harmony items combine in an amalgamation of items
from different sub-clusters within these two clusters. Thus, in order for the SAPI to be viable
in the South African context in terms of the collective dimension of personality; the
underlying structure needs to be reconceptualised and the items reworked. Particular
emphasis needs to be placed on the Integrity and Social Desirability clusters, as Integrity may
not be considered a valid measure on its own and thus needs to be rethought; while the Social
Desirability cluster MCSDS was found to relate to Integrity such that it may measure
Integrity instead of Social Desirability. However, the BIDR sub-clusters Self-deception and
Impression Management appear to be assessing something different to Integrity and the
MCSDS Social Desirability sub-cluster.
The CPAI-2, on the other hand, demonstrates adequate construct validity indicating a six
factor solution as proposed by Cheung et al. (2008). This solution was believed to make the
most sense theoretically, as Ren Qing and especially Thrift vs Extravagancevs Extravagance
were found to be problematic in the empirically imposed two factor solution. A joint factor
analysis of the SAPI and CPAI-2 scales and subscales did not replicate according to the
proposed factor structure. Thus although there is evidence of the collective dimension of
personality in the South African context, the SAPI clusters do not load as expected and the
Thrift vs Extravagance vs Extravagance scale of the CPAI-2 is not reflected in a South
176
African sample, indicating that it may be unique to Chinese culture. Thus a joint factor
analysis of the collective dimension lacks appropriate construct validity in its current form.
Construct bias was found to exist between gender and the SAPI sub-clusters- Egoism,
Hostility, Active Support, Empathy, Approachability, Harmony Maintenance and Harmony
Breach- as well as the CPAI-2 factor Interpersonal Relatedness and the scales Ren Qing,
Discipline, Social Sensitivity, Harmony. Construct bias was found for race on the dimensions
Interpersonal Relatedness, Traditionalism vs Modernity, Egoism, Hostility, Facilitating,
Guidance, Encouraging Others, Integrity, Approachability, Harmony Maintenance, Harmony
Breach and Impression Management. Lastly, construct bias for home language was revealed
for the scales dimensions Interpersonal Relatedness, Traditionalism vs Modernity,
Facilitating, Guidance, Encouraging Others, Integrity, Approachability and Impression
Management.These results may therefore suggest that bias was in fact operational in the
SAPI and CPAI-2 for this study in terms of gender, race and home language. Furthermore
these results may be indicative of more underlying differences in terms of gender, race and
home language as suggested by FFT in terms of one’s Basic Tendencies.
The quantitative results were supplemented by the qualitative thematic content analysis
performed on data collected from the focus group. The themes that emerged included a
general understanding of personality, the innate existence of personality vs the over
expression of personality, collectivism in personality, a national identity, culture,
psychometric testing and social desirability. Thus it was discovered that a unique personality
in South African is believed to exist within a national identity; inclusive of a collective
dimension indicative of a non-Western culture such as South Africa. This therefore indicates
the inadequacy of the FFT in accounting for the collective dimension of personality.
However, problems are found to exist in the expression and measurement of a collective
personality in that construct bias exists across gender, culture and language. The SAPI and
CPAI-2 were found to be long and repetitive and issues emerged with regards to the five-pint
rating scale utilised. The construct validity of the SAPI and CPAI-2 is limited as certain items
were found to linguistically and culturally inappropriate for a South African sample. Thus
these items need to be removed or reworded in order to increase its appropriateness in the
South African context, in addition to increasing its validity. Lastly, social desirability
177
emerged as an issue within personality assessment, such that the MCSDS scale used to
measure social desirability bias were found to reflect a personality trait more than a response
bias. This indicates that the Social Desirability cluster of the SAPI needs to be rethought and
refined.
A collective dimension of personality therefore can be said to have emerged from the SAPI
and CPAI-2 in a South African sample. This study utilised an etic-emic approach, allowing
these instruments to capture this dimension that is not captured in other etic measures of
personality. Although this collective dimension of personality emerged in this South African
sample, the nature of this dimension is unclear. Furthermore it is unclear whether this
collectivistic orientation is indigenous to personality in South Africa, or whether these
characteristics are universal.
178
Chapter 5: Limitations and Recommendations for Future
Research
5.1. Introduction
In this chapter the limitations of this study will be highlighted with particular reference to the
conceptual and methodological limitations. The conceptual limitations, outlined below, bring
to light the salient limitations with regard to personality assessment and the collective
dimension of personality. The methodological limitations discussed in this chapter include
limitations in terms of sample, self-report inventories and response bias, the SAPI, research
design, statistical analyses and the qualitative analysis. Following this, some
recommendations for future research areproposed.
5.2. Conceptual Limitations
The conceptual limitations inherent in personality and personality assessment as related to
this study will be presented and discussed below. Thus limitations pertaining to personality
assessment are presented first, followed by the individualism/collectivism distinction and the
nomothetic-idiographic debate. Lastly, the FFM and FFT will be discussed in terms of
collectivism.
5.2.1.Personality Assessment
The vast array of research on personality theory and assessment over the past few decades
indicates that personality is indeed measurable in some shape or form. However, Vernon
(1957) suggests that while many personality qualities can be measured quite effectively,
methods of assessment are far too elaborate, time consuming and dependent on the skill and
experience of the examiner to be applicable for any practical purpose or used by just anyone.
Freeman (1962) elaborates revealing that personality inventories attempt to measure covert
personality traits in an aim to provide a basis with which to understand individual behaviour.
179
Personality is considered to be an organisation of traits that interact. However, personality
inventories do not assess this or personality as a whole, but rather traits in isolation. Thus
personality inventories are said to be static in that they do not engage with humans as
dynamic entities (Lanyon & Goodstein, 1997).Furthermore, personality occurs at a conscious,
preconscious and unconscious level, thus in order to effectively and comprehensively
measure personality, personality inventories need to probe into these levels of awareness.
However, whether these inventories succeed in doing so is a question for debate (Freeman,
1962).
5.2.2.Individualism/Collectivism
A conceptual limitation in the current research relates to the collectivist dimension proposed
by the SAPI and CPAI-2. While there is sufficient evidence to suggest an
individualism/collectivism distinction in personality, particularly in African and Asian
cultures, this distinction is only considered to be useful in terms of analysis; as to assume
every individual of the same culture to possess the same personality characteristics is a gross
stereotyping.
Different cultures have been described as possessing individualistic and/or collectivistic
traits; this distinction is the degree to which individuals in a society are distinct or equal
(horizontal) and the degree to which individuals differ in status (vertical). Individualistic
cultures are believed to maintain their independence through expressions of their
individuality and inner attributes rather than group norms; while collectivistic cultures, such
as Asians, seek to maintain harmonious relationships assimilating in to the collective group
(Triandis, 1989; Markus &Kitayama, 1998). The collective construction of personality refers
to relationality, while individualistic constructions view and describe groups in terms of
individual identities submerged within a social identity (Markus &Kitayama, 1998; Ellemers,
Spears &Doosje, 2002; as cited in Laher, 2010). However, it has increasingly come to the
forefront that Western structures of personality are ill equipped to encompass the indigenous
traits of non-Western cultures (Costa & McCrae, 2008).
180
Studies by Cheung and colleagues have revealed the existence of a sixth factor, namely
Interpersonal Relatedness. It is believed that this factor satisfies the need for an indigenous
personality construct that are not incorporated in Western personality theory (Cheung, et al.,
2003).This collective dimension describes an indigenous cultural context that involves the
achievement of equilibrium through the qualities of humanity and social propriety; such that
individuals are found to be equal (Cheung et al., 2008).However the replication of this
collective dimension personality in a South African sample in this study revealed that this etic
measure may not be as indigenous to Chinese cultures, as originally suggested.
Within South Africa, this indigenous construct has emerged through the concept of Ubuntu.
This collective dimension of personality is described as an interpersonal element and
collective spirit that defines “the South African way”. This concept is said to possess a
collective orientation highlighted by equality and togetherness; it describes the functioning of
an individual within a community, thus emphasising the community (Louw, 2002; as cited in
Nel, 2008). It was been suggested by Nel (2008) that Bantu languages possess more
collectivistic traits; while English and Afrikaans languages possess more individualistic traits.
However, this is regarded as inadequate generalisation in that Ubuntu is common in all
language groups in South Africa, though not exclusive to them.Thus South Africans may
function in collectivistic and/or individualistic environments and thus they adopt the
necessary qualities of each environment depending on the situation (Nel, 2008).
Critics have, however, suggested that the concept of Ubuntu is overstated and does not exist
to the degree in which it was proposed. Ubuntu is said by Marx (2002; as cited in Laher,
2010) to be a political tool used by post-apartheid political elite in order to create a discourse
of ‘national identity’ and unity amongst the divided groups. Ubuntu has also been suggested
as guiding principles for the ethics within personal relationships. It is here that this study has
been espoused.Results of this study, nonetheless, reveal the existence of an indigenous
collective dimension of personality in South Africa that can be said to be representative of
elements of the concept of Ubuntu.
181
Collectivism has also been argued to be more of a narrow concept that originally suggested,
in that collectivist cultures tend to the needs of their communities. However the word
‘community’ differs amongst cultures and largely refers to the in-group or one’s family, such
that collectivism harbours individualism in a communal manner- as indicated by members of
the focus groups. Thus a debate exists as to whether the individualist/collectivist dimension
should be a separate factor or if it is a cultural mechanism that should be incorporated into the
existing items of personality inventories. This study, however, provides evidence of a
separate collective dimension in the South African context.
5.2.3. Nomothetic-Idiographic Debate
Another conceptual debate within personality is the nomothetic-idiographic debate. The
nomothetic approach aims to analyse statistical differences amongst a large sample so that it
can be generalised to a broad population. The idiographic approach, on the other hand,
attempts to gain great insights into human personality through analysing a small group of
people. This approach therefore endeavours to understand individual uniqueness (McCrae &
Costa, 2008a; Schultz & Schultz, 2008). All trait-based approaches are therefore nomothetic,
such as in the SAPI and CPAI-2; such that individual uniqueness in personality is largely
ignored. Thus the easy administration of such trait based personality inventories is offered in
contrast to the restricted information they provide.
These inventories, as well as the collective approach to personality, thus assume the
uniformity in results applied from group statistic, ignoring individual variability (England,
1991).The trait approach to personality assessment therefore limits the degree to which
instruments allow for the true exploration of emic personality constructs such as the
collective dimension in a South African context. Although the SAPI and CPAI-2 provide a
more idiographic non-Western conceptualisation of personality, both instruments subscribe to
the trait approach, with the CPAI-2 espoused in Chinese language and culture. Emic
measures are also said to experience problems in terms of bias. These problems result as emic
measures emphasise the shared experience of a culture obtained from a small minority of that
culture, such that results are not always objective.
182
5.2.4. FFM/FFT as related to Collectivism
Another theoretical limitation with regards to personality, and particularly the collectivist
dimension, is the FFM and FFT. A number of critiques have been presented for the FFM such
that it is said to be too descriptive, does not account for personality development across the
lifespan and is in fact atheoretical (McCrae& Costa, 1996).The FFT was developed by
McCrae and Costa (1996) as a result of these criticisms.
Figure 5.1: A representation of the Five Factor Theory personality system
Note. Core components are in Rectangles; interfacing components are in ellipses.
Reprinted from “Empirical and Theoretical Status of the Five Factor Model of Personality
Traits,” by R. R. McCrae and P. Costa, Jr. 2008a, The SAGE Handbook of Personality
Theory and Assessment, G. J. Boyle, G. Matthews & D. H. Saklofske (Eds.) (p.
278).Copyright 2008 by SAGE Publications, London.Reprinted with permission of the
authors.
183
McCrae and Costa (2008b) recognised the need for a unifying explanatory framework and a
theory of personality to underlie the FFM. FFT theorises the existence of a number of
components of the personality system that mutually interact to form one’s personality;
Biological Bases and External Influences are inputs into this personality system (see figure
5.1).The Objective Biography is the output of this personality system and represents
individual behaviour, thinking and feeling at any point in time (Allik& McCrae, 2004;
McCrae & Costa, 2008b). Central to the personality system is one’s Basic Tendencies and
Characteristic Adaptations. Basic Tendencies are described as traits or innate structures- the
psychological potentials of the individual rooted in biology. These traits are therefore not
susceptible to changes in the environment (Allik& McCrae, 2004; McCrae & Costa, 1996;
2008b). Characteristic Adaptations, on the other hand, are described as the structures of our
personality that an individual acquires through interaction with the environment; and thus are
susceptible to change (Allik& McCrae, 2004; McCrae & Costa, 1996; 2008b). The last
central component of this personality system is described as the Self-Concept. Theoretically
the Self-Concept forms part of Characteristic Adaptations; however is considered a separate
component. TheSelf-Concept is a view of the self, acquired as a result of one’s life
experiences and social feedback; furthermore it is shaped by one’s traits or Basic Tendencies.
Individuals tend to draw from this component when completing questionnaires such as
personality tests (Allik& McCrae, 2004).
A major criticism of the FFM deals with the number and nature of the factors that comprise
of the FFM and FFT (McCrae & Costa, 2008b). A growing body of research has emphasised
concerns as to the adequacy of five factors in comprehensively describing personality, as
postulated by the FFM and FFT (McCrae & Costa, 2008b). Specifically, however, it is
believed that this conceptualisation of personality-by virtue of its development in Western
cultures- does not address non-Western conceptualisations of personality (Yang & Bond,
1990). Studies by Cheung et al. (2003; 2008) and Ashton and Lee (2005) provide evidence of
a sixth factor of personality- commonly cited to be that of Individualism and Collectivismindicating a gap in personality theory for indigenous collective dimensions of personality that
are not incorporated in Western personality theory (Laher, 2010). This thus suggests that FFT
does not address the collectivism dimension of personality.
184
South Africans have been said to function in collectivistic and/or individualistic
environments, such that they adopt the necessary qualities of each environment depending on
the situation (Nel, 2008).All South Africans have been shown to possess a strong inclination
to be sensitive towards others and to build and maintain a community; thus revealing a
collective dimension (Nel, 2008). Ubuntu is an indigenous concept that encompasses an
interpersonal element and collective spirit of South Africans. The term Ubuntu (humaneness)
possesses a collectivistic orientation and describes a person as a person through others
(Louw, 2002; Nel, 2008).This indicates that this indigenous and collective dimension of
personality has more than likely been bypassed by the FFM. Although the concept of Ubuntu
may be linked to Extraversion and Agreeableness of the FFM, it can be argued that these
traits are assessed in an individualistic nature and so they do not capture the collectivistic
nature of this dimension of personality in South Africa.This therefore indicates that in fact
FFT may be addressing the collectivist dimension of personality; however it appears to do so
with an individualistic orientation.
Critics have argued that this concept of Ubuntu is overstated and does not exist, particularly
exclusively in the South African context (Nel, 2008). Further research about the collective
dimension of personality is still warranted. However, sufficient evidence exists to suggest a
possible distinction between individualism and collectivism in personality.By virtue of the
role of Characteristic Adaptations and External Influences within the personality system; it
can be said that a collectivistic orientations is submerged within the culturally-conditioned
phenomena of this structure. However, this collectivist dimension emerges as a product of the
interaction with one’s Basic Tendencies; while collectivism is argued by Nel (2008) and
Cheung et al. (2008) to be a separate factor. Consequently it seems that although the FFT is
addressing the collectivist dimension to a degree, it is doing so with an individualistic
orientation that is inadequate to describe the indigenous and collective element of personality
within South Africa. This may further indicate that one’s Characteristic Adaptations and
External Influences dictated by one’s culture and its subsequent norms (postulated by FFT)
are more important and influential in determining one’s personality in the South African
context than postulated by FFT.
185
5.3. Methodological Limitations
A number of methodological limitations have been identified in this study; these limitations
in this regard will be acknowledged and discussed below.
5.3.1.Self-Report Inventories and Response Bias
Self-report measures, such as the SAPI and CPAI-2, provide individual scores based on
individual reports of their own behaviour, personality style and preferences. Therefore the
risks of dishonest and socially desirable responses are present. This indicates that various
response biases may be operational for both instruments. The SAPI measures three aspects of
Social Desirability- including Social Desirability, Impression Management and Selfdeception in order to combat this. However the lack of reliability for these measures and
construct validity in this study reveal that such response bias may still be present. The CPAI2 addresses this issue by including a Social Desirability, Inconsistency and Infrequency scale
(Cheung et al., 2008). However only the Interpersonal Relatedness scale was used suggesting
that such response biases may still be present.
5.3.2.Sample
The size of the sample can be identified as a limitation of the study. Although a sample of
489 participants is considered to be adequate, it is not adequate for the number of
variablesthat were under investigation in this study. The sample size was reduced from 516
respondents as a result of incomplete questionnaires. A response rate of approximately
73.71% was reported for the study, which was found to be higher than the range of 10% and
40% reported by (Wimmer& Dominick; as cited in Laher, 2010).
In addition, the sample distribution was skewed such that groups such as gender, race and
home language were disproportionate.It is for this reason that the variables race and home
language were collapsed in order to create statistically viable group sizes for these variables.
186
The sample was obtained via a non-probability convenience sampling technique; consisting
of individuals that were easily available and willing to respond. The sample was consequently
not very representative of students in South Africa or the general population; therefore
limiting the population validity of this study. Due to the exploratory nature of this study,
nevertheless this sample is believed to be adequate.
Accordingly volunteer bias could be present influencing the results obtained in this study- it
is said that a certain kind of person is more likely to volunteer to participate in a study
(Rosenthal &Rosnow, 1991).This study explored the personality in the South African
context, thus volunteer bias could have affected the results obtained and merits further
exploration; however the collective dimension of personality was under assessment such that
no appreciable differences should have been found between respondents and nonrespondents.
In addition, psychology students were approached during lecture times in order to complete
the questionnaire at that time. Although students were told that they had the option to not
complete the questionnaire, most students stayed to do so as they felt it was the right thing to
do. This indicates that a pressure to stay may have been present (and was mentioned in the
focus groups), which may present an ethical issue, however at the same time combating
issues related to volunteer bias. Females are also said to be more likely than males to
volunteer to participate in research; this could explain the disproportionate representation of
males and females in the study, particularly in the focus groups, in addition to the
overwhelming majority of female psychology students.
The use of students in research, as mentioned earlier, is generally found to be problematic.
Student sample are used largely in psychological research, yet they are generally not
representative of the entire population. Research indicates that students responded differently
on personality inventories to adults (Allik& McCrae, 2004). However students are said to be
useful as they are a homogenous sample and have adopted the beliefs, values and practices of
their cultures. Therefore the exploration of the collective dimension of personality on these
187
students will present adequate results that would likely be even stronger in an adult sample
(McCrae, 2001).
5.3.3.The SAPI
An issue mentioned during the distribution of the questionnaire, as well as in the focus
groups, indicated that the questionnaire was too long and repetitive (particularly for the
SAPI). Participants reported feeling boredom, fatigue and discomfort during the completion
of the two instruments due to the large quantity of items. In the current study, participants
were required to complete 290 items of the SAPI and 74 items from the CPAI-2, for research
purposes. Thus no benefits could be accrued for the participants in filling out the
questionnaire. Participants of the focus group suggested that the long and repetitive
instruments may have resulted in response bias in order for participants to complete the
questionnaire efficiently.
The SAPI is a fairly new instrument, developed largely by Nel and his colleagues in 2008,
therefore is it undergoing continuous refinement and adaption. This consequently created
difficulties in this study- indicating a lack of construct validity. It is for this reason that item
bias could not be assessed, as well as the lack of information available from the developers in
terms of the original factor structure to examine bias in terms of factor analysis with a
procrustes rotation. Therefore limited information was available for this instrument- for
example the reliabilities from the normative sample for the subscales; thus limiting the
analysis of this instrument.
The SAPI theory proposed a number of scales and subscales found to represent the South
African personality; these were discussed in the methods section of this study. However it
was found in the distribution and analyses of this study that these subscales were not the
subscales originally proposed, but instead had been adapted based on the pilot study
conducted by Nel and his colleagues as they were not found to be represented in a South
African sample. Therefore the theoretical subscales were reworked based on the empirical
evidence; such thatthe “I help others to succeed” subscale contains one item, while other the
188
Social Desirability cluster and the Harmony Maintenance and Harmony Breachsub-clusters
were added to the original structure.Furthermore, problems emerged with regards to the
Integrity and Social Desirability clusters indicating that these clusters require refinement. The
Integrity sub-cluster included no sub-clusters, therefore indicating that it may not be a valid
factor on its own. Furthermore the Social Desirability sub-cluster appeared that it may be
assessing an element of Integrity rather than Social Desirability. In addition, the
Approachability sub-cluster of Relationship Harmony is defined as someone who is
approachable and accessible for others rather than placing oneself above others (Nel, 2008).
This definition therefore indicates that this sub-cluster is defined in a positive manner, however on
examination of the items in this sub-cluster it appears that they are framed in the negative thus
describing an individual who is not approachable or accessible. Therefore, the definition of this subcluster needs to be reformed.
5.3.4.Research design
The research design in this study can be classified as a non-experimental design, as there was
no randomisation, no assigned control group, and no variables were manipulated. Nonexperimental research designs have several disadvantages; one being that the absence of
randomisationwhich allows for possible extraneous variables. Furthermore, as a result of no
variable manipulation, this design prohibits causal interpretation.However, non-experimental
research designs allow for the discovery of trends, which are consistent with the exploratory
nature of this study (Rosenthal &Rosnow, 1991).
5.3.5.Statistical Analysis
This research was limited in the procedure undertaken to explore reliability of the SAPI and
CPAI-2 in that only internal consistency reliability was explored. Test-retest reliability, as
well as alternate form reliability were not examined in this study as they were beyond the
scope of the study; however these reliability analyses should be considered for future
research. In terms of validity, only construct validity was explored. The value of face validity
is debated in psychometric research; however as two of the research questions aimed to
ascertain the students perceptions of personality and the two assessments under investigation
this analysis would have added value to this research. Students in the focus groups did allude
189
to face validity in that they questioned the inclusion of a number of items, such as “I eat
snacks sometimes”, in their applicability in terms of personality. Content and criterion
validity were not explored as they were beyond the scope of this study, however should be
considered for future research.
Bias refers to systematic differences in measurement or prediction across groups (Murphy
&Davidshofer, 2005). A difficulty exists in South Africa with regards to how to decide on the
constitution of these groups, such that it needs to be standardised to allow for cross-cultural
validity. A limitation may also exist in the use of the word ‘race’, yet this issue is still
debatable. In terms of bias, this study only explored construct bias, and not method bias and
item bias. This should thus be considered for future research. Item bias was proposed to be
explored originally however, as expressed above, item bias could not be undertaken due to
the lack of available data from the normative study of the SAPI. The construct validity
investigation of the SAPI revealed a number of issues with regard to the conceptualisation of
the model as well as its expression of the South African personality; therefore an exploration
of item bias would not add much more value to this exploration. Furthermore, Cliff’s d could
not be calculated in this study for the non-parametric significant Kruskal-Wallis results, as
problems were encountered with regards to the formulas used to calculate this analysis.
Therefore this should be considered in future research.
Although factor analysis is the method of choice for validation studies within personality
psychology, it is not without its drawbacks (Kline, 1993; Murphy &Davidshofer, 2005). One
of these drawbacks, it is argued, includes the inability of this technique to determine which
factor solution is better; this is at the discretion of the researcher and is therefore subjective. It
can therefore be argued that factor analysis can be valuable or redundant depending on the
context in which you are using it. However, value of this technique is proposed for
exploratory factor analysis when the number of factors and the rotation techniques selected
based on theoretical and empirical considerations (Hopwood &Donnellan, 2010; as cited in
Laher, 2010).
190
In the current study factor analyses were conducted on a heterogeneous sample of 489
students; this sample size is adequate. Due to the ANOVA results, there is evidence of
differences across gender, race and home language; therefore a single analysis across groups
may be problematic. Separate factor analyses were not conducted on the collapsed groups
which provides a limitation to the study. Therefore a factor analysis needs to be conducted for
each group with a larger and more representative sample.
5.3.6.Qualitative Analysis
Through a qualitative exploration the researcher allowed participants a voice, which permits
the carving out of unacknowledged pieces of narrative from which evidence is selected,
edited, and deployed to border the researcher’s arguments (Fine, 2002).
Thematic content analysis is a method for identifying, analysing and reporting themes within
data; minimally organising and describing the data set in detail, as well as interpreting it
(Boyatzis, 1998; Braun & Clarke, 2006). Themes emerge during the process of analysis
however the researcher plays an active role- identifying, selecting and reporting themes
(Braun & Clarke, 2006;Taylor & Ussher, 2001). The researcher possesses the autonomy to
determine the themes and the information of interest.
In order to ensure the reliability and validity of qualitative research, the researcher needs to
assess the credibility, transferability, dependability and conformability of the qualitative
analysis.These criteria are used to judge the trustworthiness of the qualitative research
(Lincoln &Guba, 1985; as cited in Rumrill, Cook & Wiley, 2011). These standards were
designed to facilitate researcher reflexivity throughout the research process.Reflexivity is
therefore used as a methodological tool to better represent, legitimise, or call into question a
scholar’s data, such that information that supports the researcher’s argument can be selected
and edited (Fine, 2002; Pillow, 2003).
191
Researchers therefore need to be cognisant of their role in the study and how they influence
the manner in which participants communicate with them. Researchers must also be aware of
their own biases and how this may affect misinterpretations of data (Creswell, 2009), and to
make use of techniques that will enhance the trustworthiness of the research.Trustworthiness
can be achieved through numerous techniques that include, audit trails, field notes, memos,
triangulations, discrepant data, member checks and peer debriefs (Lincoln &Guba, 1985; as
cited in Rumrill, Cook & Wiley, 2011).
In this study, I continuously made attempts to self-reflect on my role as a research and to
remain as neutral as possible. This, however, would have been more effective had I
documented my feelings and thoughts in a journal throughout the research process.The
thematic content analysis process was documented throughout, thus providing a type of audit
trail of raw data, video tapes, focus group notes,process notes, products of coding procedures
and products of data reconstruction; thus providing a means for dependability and
conformability.Participants of the focus group, who also participated in the study, were asked
to explore the two instruments and discuss personality and personality assessment in the
South African context; providing a type of member check for the research.Furthermore
within focus group two results of the first focus group were discussed when relevant, thus
assessing credibility of the research to a degree. My ideas and interpretations for the focus
groups were clarified by my supervisor, therefore achieving a kind of peer debrief. Within the
focus group sessions reactivity was managed by the researcher by using open-ended
statements, reflective statements and the restatement of participants’ comments. However it is
impossible to completely prevent reactivity in a focus group where the researcher actively
engages with participants (Rumrill, Cook & Wiley, 2011).Therefore, it can be established that
attempts were made throughout the research process to control for research bias and
reactivity; however the trustworthiness of the research, and consequently the reliability and
validity, can be improved by making use of various other trustworthiness techniques as
described above.
192
5.4. Recommendations for Future Research
Based on the literature review, the results of this study and the limitations presented above,
the following recommendations for future research with regards to the collective dimension
of personality in South Africa is presented below.
More in depth studies utilising a mixed methods approach need to be conducted in order to
obtain rich data with regards to emic measures of personality within the South African
context. This calls for more research into the Individualism/Collectivism in South Africa in
order to ass to the etic-emic debate locally and internationally. This will allow for a greater
understanding of the collective dimension of personality and particularly personality in South
Africa.
Further research is required for the Interpersonal Relatedness subscale of the CPAI-2 in the
South African context in order to further understand the relationship between these collective
dimension variables and personality in South Africa; particularly in terms of Harmony,
Traditionalism vs Modernity and the problematic subscalesRen Qing and Thrift vs
Extravagance. Therefore items with regards to these dimensions need to be adapted for the
South African context, as they do not measure personality in South Africa as they stand.
In terms of the SAPI- the collective dimensions proposed in this model need to be
reconceptualised in order to appropriately group the dimensions. Items need to be revised and
adapted according to the results found, such that items are reconceptualised to form other
constructs, as suggested by this study, or removed. Following from which, further analysis in
the South African context is required in order to obtain the collective dimensions that reflect
personality in this context. Longitudinal studies may allow for the establishment of trends, as
well as test-retest reliability; such that personality can be explored over time in order to
examine the stability of personality and the comparison with international trends.
193
A study testing the reliability and validity of the FFM in South Africa, administered the
NEO-PI-R to 368 students from four different universities, found that the participant’s
personality traits varied as a function of race, educational level and socio-economic status
(Zhang &Akande, 2002). Thus socio-economic status may be interesting variable to explore
in addition to gender, race and home language; as issues with regard to socio-economic status
is present in a large majority of South African lives such that it may have a role in personality
development.
Similarly, acculturation is argued as an impeding issue in the expression and understanding of
personality, such that acculturation measures should be included in future studies.
Acculturation, as discussed in Chapter 4, argues the modification of original cultural patterns
of different groups as a result of their continuous direct contact with other cultures (Heuchert
et al., 2000). Participants of the focus groups in this study indicated that the new generation
within South Africa is becoming a mix of individualism and collectivism as a result of our
adoption of ‘Western’ ideals of individualism, as well as the belief in the ‘spirit of the
community’- an ideal of collectivism. This argument highlights that proposed by
acculturation presented by Okeke et al. (1999; as cited in Laher, 2010) and Heuchert et al.
(2000). Acculturation,thus, bring into question the existence of cultural differences, or
alternatively, merely the degree to which these differences are expressed (Eaton &Louw,
2002). While studies reveal that overt behaviours conform to that of the dominant culture, but
with the existence of hidden traditional cultural elements (Mpofu, 2001); Berry and Sam
(1997) propose a bidimensional model of acculturation. This model depicts four types of
acculturation- integration, separation, assimilation and marginalisation- that depend on the
degree to which one maintains a good relationship with their host culture or native culture
(Berry & Sam, 1997). Thus, due to the composition of the sample in this study it is
imperative that this argument of acculturation be addressed in future studies.
A larger and more representative sample of South Africans utilising stratified samples would
be recommended for future research. This will ensure greater generalisability of results and
greater adequacy of statistical analyses. Additionally, since volunteer bias is problematic,
future studies should attempt to include non-volunteer participants in order to combat this
194
issue. The sample should also be expanded to include other participants in addition to
students.
More emphasis should also be placed on the effects of social desirability and response bias in
psychometric testing of personality- particularly for use in organisational settings. Response
bias is a source of error that describes the tendency of individuals to answer in such a way
that does not reflect their true feelings, beliefs and behaviours. Response bias can therefore
result in an underestimation or overestimation of variables (Bradburn, 1983; as cited in
Monette, Sullivan&DeJong, 2010).Types of response bias include acquiescence response set,
response pattern anxiety and social desirability. Therefore in order to reduce systematic or
random error, these sources of response bias must be reduced (Monette et al.,
2010).Acquiescence bias is said to be a result of cultural difference, such that collectivistic
cultures tend to be more acquiescent that individualistic cultures (Smith, 2004; as cited in
Leite&Beretvas, 2005).
Social desirability is of particular interest in this study and hence needs to be controlled for.
In completing psychometric tests individuals attempt to display their best qualities, and often
the qualities that the organisation is looking for, in order to be selected for a job;thus
indicating the influence that social desirability may have in selection decisions. In such
selection proceedings, amongst others, where individuals are required to make a positive
impression social desirability has been identified as a well-known dilemma (Levin &Montag,
1987; as cited in Ones et al., 1996). It appears from the analyses that the Social Desirability
cluster of the SAPI experienced conceptualisation and scale construction issues; thus bringing
into question the extent at which this response bias was controlled for. A frequent criticism of
social desirability dictates that it “is a source of bias only to the extent that it represents
respondents’ conscious deception efforts” (Uziel, 2010, p. 244); thus indicating that those
socially desirable behaviours accessible to the individual’s conscious mind will be assessed
by this scale while it is doubtful whether individual unconscious defences are assessable.It is
also unclear whether social desirability in fact forms its own personality variable or if it is a
component of various personality dimensions (Weiner &Craighead, 2010).Rampant debates
over the past few decades have called into question whether social desirability scales as a
reliable validity measure or one that detects bias reporting (Holden, 2008; Piedmont et al.,
195
2000; Uziel, 2010). Although research provides evidence to support this view, test
responding issues such as social desirability are considered a valuable consideration in
personality assessments (Weiner &Craighead, 2010).
Further research should be done comparing responses on the SAPI and CPAI-2 against other
prominent personality instruments, such as the 16PF, OPQ and particularly the BTI, as it is a
South African developed instrument, in order to consider criterion validity. This will also
allow for further discussion on the etic-emic debate with regards to personality assessment in
South Africa. Theoretically this will allow further insight into the utility of the SAPI and
CPAI-2 models in representing the collective dimension of personality in South Africa.
5.5. Conclusion
This study provided evidence of a collective dimension of personality in South Africa as
measured by the two instruments. However, the factor structure proposed in terms of this
collective dimension in these instruments did not replicate entirely in a South African sample.
Evidence of the CPAI-2 Interpersonal Relatedness factors indicated a collective dimension in
South Africa, therefore suggesting that this factor may not be unique to Chinese culture.
However the subscales Ren Qing and Thrift vs Extravagance were not found in this South
African sample thus indicating that these factors may not be applicable to personality in
South Africa or may be applicable in a different form or expression.
In terms of the SAPI, evidence for all four collective dimensions- Facilitating, SoftHeartedness, Relationship Harmony, Integrity - and Social Desirability- were found; however
the structure did not replicate as expected, with items from each scale loading on different
factors. The Integrity and Social Desirability scales were found to be particularly
problematic. Therefore the results suggested that emic instruments do have value in the South
African context but need further refinement and research.
196
Systematic differences across gender, race and home language were found in this sample for
the two instruments; thus suggesting the existence of bias that needs to be explored,
particularly for the Traditionalism vs Modernity, Facilitating and Guidance dimensions.
This study has provided an initial exploration into the collective dimension of personality in
South Africa, indicating a number of findings and consequent research possibilities for
personality and personality assessment in South Africa. Further exploration into the issues
identified is necessary in order to determine whether the SAPI and the CPAI-2 are
appropriate to capture the collective dimension of personality in South Africa, and whether
the adaptations suggested are in fact feasible. Such research is also necessary so as to explore
implications for the incorporation of indigenous constructs into personality theory locally and
abroad.
Exploring the collective dimension of personality using the SAPI
and CPAI-2 in a sample of South African students.
Leah Branco e Silva
A dissertation submitted to the Faculty of Humanities, University of the Witwatersrand,
Johannesburg in partial fulfillment of the requirements for the Degree of Masters in
Organisational Psychology.
February 2012
Abstract
The Five-Factor Model (FFM) is one of the most dominant and widely acceptedframeworks of
personality, however it has been criticised for its primarily Eurocentric individualistic focus on
aspects of personality as a result of its development within the English lexicon (McCrae
&Terracciano, 2005).The use of imported etic inventories presents an array of difficulties for
personality assessment within South Africa, as indigenous and collective personality traits are
absent from these measures of personality (Nel, 2008). This study, therefore, aimed to explore
the collective dimension of personality using the South African Personality Inventory (SAPI)
and the Cross-cultural Personality Assessment Inventory-2 (CPAI-2) in a non-probability
convenience sample of 489 South African first-year Psychology students from the University
of the Witwatersrand.
A quantitative section was conducted to investigate the internal consistency reliability,
construct validity and construct biasof the collective dimension of the two instruments. From
the results it was evident that both instruments had adequate internal consistency reliability for
the scales and subscales of the collective dimensions of personality, but some reliability
coefficients were low; particularly for Social Desirability and its sub-clusters, as well as Ren
Qing, Discipline and Thrift vs Extravagance. Construct validity for the SAPI was found to be
problematic in that the five factor structure proposed by Nel (2008) did not replicate; however
a three factor structure was found to be the best fit. Construct validity for the CPAI-2 was
found to be good in that the six factor structure of the Interpersonal Relatedness factor
proposed by Cheung et al. (2008) was replicated. Evidence for construct bias across gender,
race and home language was found in both instruments. Finally a qualitative thematic content
analysis of data obtained from two focus groups of 19 first year Psychology students from the
original sample was conducted and issues related to a general understanding of personality,
the innate existence of personality vs the overt expression of personality, collectivism in
personality, a national identity, culture, psychometric testing and social desirability were
explored. This study thus contributed to both the local and international context in terms of
personality theory and assessment in organisations, as well as theoretically to the etic-emic
debate.
Declaration
I declare that this dissertation is my own, unaided work. It is being submitted for the degree of
Masters in Organisational Psychology at the University of the Witwatersrand, Johannesburg. It
has not been submitted before for any degree or examination at any other university.
__________________
L. M. Branco e Silva
____ day of ________, 2012
Acknowledgements
This has been a very long and tedious journey, and it feels so glorious to finally put my thesis
to rest.
I would like to thank my supervisor Dr SumayaLaher for all her time, effort and dedication in
helping me complete this dissertation; without you this would not have been possible. Thank
you for always taking such care with my work and for your constant support and
encouragement. Your work ethic is inspiring and it has been a great pleasure working with
you over the last two years.
To the Wits Masters class of 2011, thank you for being a pillar of strength and support when I
felt like throwing in the towel. The high standard provided by every member of the class
forced me to work harder and aim higher to achieve my best. A special thanks to Cara and
Dani for always being there for me and pushing me when times got rough; soldiering through
the last year with you has been an adventure.
To my family and friends, thank you for putting up with my incessant bad moods and
complaining over the last year. I know it has been difficult to be around me at times, but I am
glad to say that it is all behind us! Your love, care, encouragement and support is everything to
me. I love you lots!
Table of Contents
Abstract…………………………………………………………………………..i
Declaration…………………………………………………………………...…ii
Acknowledgements………………………………………………………….…iii
Table of Contents………………………………………………………………iv
List of Tables ………………...…………………………………………….….xii
List of Figures…………………………………………………………………xiv
Chapter 1: Literature Review……………………………………...………......1
1.1. Introduction……………...…………………………………………...……...1
1.2. The Trait Approach to Personality………………………..…………...…….6
1.3. The Five-Factor Model …………………………………………………......7
1.3.1. The Five Factors……………………………………………………...7
1.3.2. Cross-Cultural Applicability of the FFM…………………………….8
1.3.3. Five or More Factors?........................................................................10
1.3.4. The FFM in South Africa…………………………………………...10
1.3.5. Problems with the FFM……………………………………………..11
1.4.Individualism vs Collectivism…………………………………………......12
1.5. The Etic-Emic Debate…………………………………………………...…15
1.6. The South African Personality Inventory……………………….………....17
1.7. The Cross-cultural Personality Assessment Inventory-2………………......19
1.8. Examining the Reliability of the SAPI and CPAI-2 Collective Scales…….22
1.9. Examining the Validity of the SAPI and CPAI-2 Collective Scales……....23
1.10. Examining Bias in the SAPI and CPAI-2 Collective Scales……………..23
1.10.1. Race and Personality………………………………………………25
1.10.2. Home Language and Personality………………………………….25
1.10.3. Gender and Personality……………………………………………26
1.11. Conclusion………………………………………………………………27
Chapter 2: Methods…………………………………………………………...28
2.1. Aims of the Study……………………………………………………….....28
2.2. Rationale for the Study………………………………………………….....28
2.3. Research Questions………………………………………………………...31
2.3.1. Quantitative Phase………………………………………………….31
2.3.2. Hypotheses………………………………………………………….32
2.3.3. Qualitative Phase……………………………………………………32
2.4. Sample……………………………………………………………………...32
2.4.1.Quantitative Phase ….………………………………………………32
2.4.2.Qualitative Phase …………………………………………………...36
2.5. Instruments…………………………………………………………………36
2.5.1. Quantitative Phase………………………………………………….36
2.5.1.1. Demographic section…...………………………………………37
2.5.1.2. The SAPI…………………………...…………………………..37
2.5.1.3. The CPAI-2……………………………………...……………..39
2.5.2. Qualitative Phase……………………………………………………41
2.6. Procedure…...……………………………………………………………...41
2.6.1.Quantitative Phase…………………………………………………..41
2.6.2.Qualitative Phase……………………………………………………42
2.7. Research Design…………………...……………………………………….43
2.7.1. Quantitative Design…………………………………………………43
2.7.2. Qualitative Design…………………………………………………..43
2.8. Ethical Considerations……………………………………………………..45
2.8.1. Quantitative Phase...………………………………………………..45
2.8.2.Qualitative Phase….………………………………………………..45
2.9. Data Analysis………………………………………………………………46
2.9.1.Quantitative Phase…………………………………………………..46
2.9.1.1. Internal Consistency Reliability………………………………..46
2.9.1.2. Construct Validity…………………………………...…………47
2.9.1.3. Construct Bias……………...…………………………………..47
2.9.2. Qualitative Phase…………………………………………...………48
2.10. Self-Reflexivity…………………………………………………………...48
2.11. Conclusion………………………………………………………………..49
Chapter 3: Results……………………………………………………………..50
3.1. Introduction…………………………………………………………….......50
3.2.Descriptive Statistics: Demographic Information……………….………....50
3.3.Descriptive Statistics: SAPI………………………………………………..51
3.4. Descriptive Statistics: CPAI-2……………………………………………..53
3.5. Reliability: SAPI and CPAI-2……………………………………………...55
3.5.1. Internal Consistency Reliability: SAPI and CPAI-2………………..55
3.6. Construct Validity- SAPI and CPAI-2……………………………………..56
3.6.1. Factor Analysis- SAPI……………………………………………...57
3.6.1.1. Scale Level……………………………………………………..57
3.6.1.1.1. Five Factor Solution………………………………..…...57
3.6.1.1.2. Three Factor Solution…………………….……….…….60
3.6.1.1.3. Four Factor Solution………………………...…………..61
3.6.1.1.4. Two Factor Solution…..………………………………...63
3.6.1.2. Item Level……………………...………………………………64
3.6.1.2.1. Five Factor Solution……...……………………………...65
3.6.1.2.2. Fourteen Factor Solution……………………………...…67
3.6.2. Factor Analysis- CPAI-2 Scale level……………………………….69
3.6.2.1. Six Factor Solution……………………………………………..69
3.6.2.2. Two Factor Solution……………………………………...…….71
3.6.3. Factor Analysis SAPI and CPAI-2………………………………….72
3.6.3.1. Subscales…………………………………...…………………..72
3.6.3.2. Scales…………………………………………………...……...75
3.6.3.3. Six Factor Solution……………………………………..……...77
3.7. Construct Bias for the SAPI and CPAI-2…………………………………..78
3.7.1. Gender- The SAPI and CPAI-2…………………………………….78
3.7.2. Race- The SAPI and CPAI-2……………………………………….80
3.7.3. Home Language- The SAPI and CPAI-2…………………………...82
3.8. Qualitative Data: Thematic Content Analysis……………………………..84
3.8.1. General Understandings of Personality…………………………….85
3.8.2. The Innate Existence of Personality vs the Overt Expression of
Personality……………………..…………………...………………86
3.8.3. Collectivism in Personality…………………………………………87
3.8.3.1. Urban vs Rural………...……………………………………….88
3.8.4. A National Identity………………………………………………….89
3.8.5. Culture………………………………………………………………92
3.8.6. Psychometric Testing……………………………………………….94
3.8.6.1. Length…………...……………………………………………..94
3.8.6.2. Repetition………………………..……………………………..94
3.8.6.3. Rating Scale……………………...…………………………….95
3.8.6.4. Problematic Items…………………………...…………………96
3.8.6.4.1. Linguistically Problematic Items………………………...96
3.8.6.4.2. Culturally Problematic Items…………………………….98
3.8.7. Social Desirability…………………………………………………100
3.9. Conclusion………………………………………………………………..101
Chapter 4: Discussion………………………………………………………..104
4.1. Introduction……………………………………………………………….104
4.2. Descriptive Statistic: Demographic Information…………………………104
4.3. Descriptive Statistics: SAPI………………………………………………106
4.4. Descriptive Statistics: CPAI-2……………………………………………106
4.5. Internal Consistency Reliability: SAPI…………………………………...107
4.6. Internal Consistency Reliability: CPAI-2………………………………...111
4.7. Construct Validity: SAPI…………………………………………………113
4.7.1. Scale Level………………………………………………………...113
4.7.1.1. Factor Solution including Social Desirability..…………...…..115
4.7.1.2. Factor Solution excluding Social Desirability…...…………...120
4.7.2. Item Level: SAPI………………………………………………….125
4.7.2.1. Five and Fourteen Factor Solution……………………...…….130
4.8. Construct Validity: CPAI-2………………………………………………138
4.8.1. Scale Level………………………………………………………...138
4.8.1.1. Six factor solution……...……………………………………..139
4.8.1.2. Two Factor Solution…………………………………………..143
4.9. Construct Validity: SAPI and CPAI-2……………………………………145
4.9.1. Subscales………………….……………………………………….145
4.9.2. Scales……………………………………………………………...149
4.9.3. Six Factor Solution………………………………………………...150
4.10. Construct Bias: SAPI and CPAI-2………………………………………152
4.10.1. Construct Bias across Gender……………………………………152
4.10.2. Construct Bias across Race………………………………………154
4.10.3. Construct Bias across Home Language..………………………...157
4.11. Thematic Content Analysis……………………………………………...159
4.11.1. General Understandings of Personality………………………….159
4.11.2. The Innate Existence of Personality vs the Overt Expression of
Personality………………………………………...…………….160
4.11.3. Collectivism in Personality………………………...……………162
4.11.4. A National Identity……………………...……………………….164
4.11.5. Culture…………………...………………………………………165
4.11.6. Psychometric Testing………………………...………………….167
4.11.7. Social Desirability…...…………………………………………..171
4.12. Conclusion………………………………………………………………174
Chapter 5: Limitations and recommendations for future research……....178
5.1. Introduction……………………………………………………………….178
5.2. Conceptual Limitations…………………………………………………...178
5.2.1. Personality Assessment………...………………………………….178
5.2.2. Individualism/Collectivism………………………………………..179
5.2.3. Nomothetic-Idiographic Debate………...…………………………181
5.2.4. FFM/FFT as related to Collectivism………………………………182
5.3. Methodological Limitations………………………………………………185
5.3.1. Self-Report Inventories and Response Bias……………………….185
5.3.2. Sample……………………………………………………………..185
5.3.3. The SAPI…………………………………………………………..187
5.3.4. Research design……………………………………………………188
5.3.5. Statistical Analysis………………………………………………...188
5.3.6. Qualitative Analysis……………………………………………….190
5.4. Recommendations for Future Research…………………………………..192
5.5. Conclusion………………………………………………………………..195
Reference List………………………………………………………………..197
Appendix A- Sample Questionnaire………………………………………..211
Participant Information Sheet (Quantitative phase)………………………..211
Questionnaire directions……………………………………………………212
SH1: Sample SAPI items…………………………………………………...213
SH2: Sample CPAI-2 items………………………………………………...215
SAPI Answer Sheet…………………………………………………………217
Appendix B- Focus Group Participation…………………………………...218
Participant Information Sheet (Qualitative phase)……………………….....218
Informed Consent Form (Focus Group)………………………………….....219
Informed Consent Form (Recording)…………………………………........220
Appendix C- Focus Group Questions...………………………………….....221
List of Tables
Table 2.1: Demographic statistics for the sample…………………………………………34
Table 2.2: Demographic statistics for age and English reading ability………………….35
Table 2.3: Descriptive statistics for gender and race……………………………………..36
Table 3.1: Frequencies for gender, race and home language…………………………….51
Table 3.2: Descriptive Statistics for the SAPI…………………………………………….52
Table 3.3: Descriptive Statistics for the CPAI-2………………………………………….54
Table 3.4: Internal consistency reliability coefficients for the SAPI and CPAI-2………55
Table 3.5: Eigenvalues for the SAPI five factor solution…………………………………58
Table 3.6: Five factor solution for the SAPI using varimax rotation……………………59
Table 3.7: Three factor solution for the SAPI using varimax rotation………………….60
Table 3.8: Eigenvalues for the SAPI four factor solution………………………………...61
Table 3.9: Four factor solution for the SAPI using varimax rotation…………..……….62
Table 3.10: Two factor solution for the SAPI using varimax rotation…………………..63
Table 3.11:Eigenvalues for the SAPI at item level……………………………………….64
Table 3.12:Eigenvalues for the CPAI-2 scales……………………………………………70
Table 3.13: Six factor solution for the CPAI-2 using varimax rotation…………………71
Table 3.14: Two factor solution for the CPAI-2 using varimax rotation………………..72
Table 3.15:Eigenvalues for the SAPI and CPAI-2 subscales……………………………73
Table 3.16: Factor solution for the SAPI and CPAI-2 subscales usingVarimax
rotation……………………………………………………………...................74
Table 3.17:Eigenvalues for the SAPI and CPAI-2 scales………………………………..75
Table 3.18: Factor solution for the SAPI and CPAI-2 scales using varimax
Rotation…………………………………………………………………………76
Table 3.19: Six factor solution for the SAPI and CPAI-2 using varimax rotation…….77
Table 3.20: ANOVA results for gender on the SAPI and CPAI-2………………………79
Table 3.21: ANOVA results for race on the SAPI and CPAI-2………………………….81
Table 3.22: ANOVA results for home language on the SAPI and CPAI-2……………..83
Table 4.1: Cronbach Alpha coefficients for the SAPI scales……………………………110
Table 4.2: Cronbach Alpha coefficients for the CAPI-2 scales…………………………111
Table 4.3: Five and Fourteen Factor solution results for the SAPI and CPAI-2
Items using varimaxrotation………………………………………………....126
List of Figures
Figure 3.1: Cattell’s Scree Plot for SAPI Subclusters five factor solution………...…….58
Figure 3.2: Cattell’s Scree Plot for SAPI Subclusters four factor solution……….……..62
Figure 3.3: Cattell’s Scree Plot for SAPI at item level…………………….……………...65
Figure 3.4: Cattell’s Scree Plot for CPAI-2 Scales…………………………….………….70
Figure 3.5: Cattell’s Scree Plot for SAPI and CPAI-2 subscales………….……………..73
Figure 3.6: Cattell’s Scree Plot for SAPI and CPAI-2 Scales…………………………....76
Figure 5.1: A representation of the Five Factor Theory personality system…………..182
Psychology
School of Human & Community Development
University of the Witwatersrand
Private Bag 3, WITS, 2050
Tel: (011) 717 4500
Fax: (011) 717 4559
Dear Student
Hi! We are currently conducting research on the interpersonal relatedness factors of personality. This
study forms part of a larger study on personality in South Africa. Part of this research requires your
responses on the attached questionnaire. It should take you approximately 1½ hours to complete the
questionnaire. Please ensure that you answer on the answer sheet. I understand that this is a
substantial investment of your time. However your response is valuable as it will contribute towards a
South African understanding of personality and will have an impact on research nationally and
internationally. I would therefore like to invite you to participate in this research.
Your responses will remain confidential and anonymity is guaranteed. At no time will I know who you are
since the questionnaire requires no identifying information. Completion and return of the questionnaire will
be considered to indicate permission for me to use your responses for the research project. Should you
choose not to participate, this will not be held against you in any way. As I am only interested in group
trends, and have no way of linking any individual’s identity to a particular questionnaire, I will not be able
to give you individual feedback. Feedback in the form of a one-page summary sheet will be available on
request approximately six months after completion of this questionnaire. If you have any further questions
or require feedback on the progress of the research, please feel free to contact me. My contact details
appear below my signature. The data from this study will be used in conjunction with data collected from
other South African samples to determine the value of the South African Personality Inventory.
Thank you for considering taking part in the research project. Please detach and keep this sheet.
__________________
Dr. S. Laher
011 717 4532 / [email protected]
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
SH 1
DIRECTIONS
This booklet contains statements about your interests, preferences, everyday thoughts and
behaviours and how you feel about things. You need to respond to all statements in the
booklet. The questionnaire contains 364 statements. Please use an HB pencil to complete the
questionnaire. Darken the circle that corresponds to the number that best indicates your
answer:
1. Strongly Disagree (1)
2. Disagree (2)
3. Somewhat Disagree / Somewhat Agree (3)
4. Agree (4)
5. Strongly Agree (5)
When the administrator asks you to start, answer each statement. Please keep the following
things in mind:
1. Read each statement and choose the answer that best describes you. There are no “right”
or “wrong” answers; just mark what is true for you.
2. If you want to change an answer, please erase the original one completely and darken the
circle of the number that best indicates your answer.
3. Please answer truthfully.
4. Complete the biographical information.
Please ask now if you have any questions.
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
SH 1
INSTRUCTIONS PART 1
Please read each of the statements below and indicate your disagreement or agreement by
darkening the circle on your answer sheet that best indicates your answer:
1.
I help people grow
2.
I am loyal to others
3.
I greet others
4.
I accept change
5.
I boast about things that make me
better than others
6.
I accept others
7.
I am suspicious of what people close to
me do
8.
I tell the truth
9.
I help people find their way in life
10.
I listen to other people's problems
11.
I am a nice person
12.
I challenge people in front of others
13.
I support others when they have to
make difficult decisions
hurt me
22.
I want to sort things out in a peaceful
manner
23.
I provide guidance to help others
succeed
24.
I give others emotional support
25.
I interfere in the lives of others
26.
I look after other people’s belongings
when they are away
27.
I maintain relations with others for a
long time
28.
I think about my options before I make
a choice
29.
I play a leading role in the lives of
people close to me
30.
I have a low opinion of others
31.
When I hear people talking privately, I
avoid listening
I give advice that people would keep
for a long time
32.
I hurt others
15.
I spread rumours about others
33.
I teach people ways of doing things
16.
I hide from others who I really am
34.
I cause arguments between others
17.
I criticise others' mistakes
35.
I remain serious when jokes are told
18.
I sometimes regret my decisions
36.
I keep my things for myself
19.
I protect others
37.
I forgive easily
20.
I fake a smile when I have to
38.
I wish others to be successful
21.
I take revenge on people who have
39.
I uplift others
14.
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
40.
I behave differently from how I really
feel
41.
I care for others like for my own family
42.
I have disliked it when others are
happy
43.
I regard my background as being better
than the backgrounds of others
44.
I have some bad habits
45.
I have done things that I do not tell
other people about
46.
I pass my passion about things on to
others
47.
I treat weaker people badly
48.
I take others’ feelings into account
49.
I talk to people in a respectful way
50.
I make others feel comfortable
51.
Once I have made up my mind, other
people cannot change my opinion
52.
I show people how to live their life in a
good way
53.
I can use many different ways of doing
things
54.
I give people directions in life
55.
I criticise people when I do not
understand them
56.
I continue with my work if I am
motivated
67.
I take advantage of weaker people
68.
I have hurt others with my words
69.
I sometimes tell lies if I have to
70.
I believe that there is nothing to learn
from other people
71.
I am approachable when others need
help
72.
I am a person everybody listens to
73.
I discriminate against people
74.
I have taken things from others using
force
75.
It is hard for me to break my bad habits
76.
I say bad things about others while
they are not there
77.
I only think of myself
78.
I am open to people with different
opinions
79.
I abuse my power over others
80.
I have humiliated others
81.
I provide directions to others about
what is right and wrong
82.
I like doing things for others
83.
I live by my principles
84.
I have doubts about my ability to
succeed in life
85.
I support people close to me in
whatever they choose to do
86.
I lead people by showing good
behaviour
57.
I make people do things for me
58.
I am just an average person
59.
I have done things that are against the
law
87.
I keep people's attention when I explain
something
60.
I am happy when I hear bad news
about people whom I do not like
88.
I accept apologies
61.
I tell people how they can cope with
difficulties
89.
I am very confident of my judgments
90.
I respect others’ opinions
62.
I start fights with others
91.
I care about other people's future
63.
I advise others when they are in
unfamiliar situations
92.
I provoke others
93.
I am a good leader
64.
I get involved in other's affairs
94.
I have been violent in the past
65.
I accept the advice of others
95.
I resolve conflicts between people
66.
I guide people in life
96.
I help others when they are in need
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
97.
different subjects
I believe that many people have bad
intentions
128.
I am kind to others
98.
I am easy to talk to
129.
I keep my promises
99.
I can see others’ point of view
130.
I avoid arguments
100.
I create tension between others
131.
101.
I am sensitive to other people’s feelings
I make people feel involved when I
explain something
102.
I disappoint others
132.
I make jokes about other people
133.
I make people feel weak when they are
around me
134.
I can share in someone's emotions
135.
I choose the people I want to speak to
136.
I allow others to ask me questions
137.
I have frightened others
138.
I make others feel at home
139.
I have so much pride that I find it
difficult to accept others
103.
My first impressions of people usually
turn out to be right
104.
I hug people when I am happy about
them
105.
I want people to live in peace
106.
I look down on others
107.
I indulge others
108.
I value others for what they are
109.
I do what is expected of me
140.
I gossip
110.
I think I am more important than others
141.
I help people to solve their arguments
111.
I have embarrassed people in front of
others
142.
I find ways to help people even if it is
difficult
112.
I am careful about my way of dressing
143.
113.
I show humanity towards others
I influence people in important ways in
their lives
114.
I have a clean mind about my actions
144.
I am a true friend to others
115.
I treat others in a careful way
145.
I ask people if they are all right
116.
I only see the bad sides of people
146.
I encourage others when they face
difficulties
117.
I favour some people above others
147.
I am friendly towards others
118.
I find it difficult to confide in others
148.
I stand by what I say
119.
I treat some people badly because of
their background
149.
I show people the right way when they
are unsure about something
120.
I am an example for others
150.
121.
I give things to people without
expecting anything in return
I discuss people's personal affairs with
others
151.
I look after the safety of others
122.
I do the right thing
152.
I am a source of inspiration to people
123.
I listen carefully to understand what
others are saying
153.
I come home at the expected time
124.
I take responsibility for my mistakes
154.
I give my attention to others
125.
I want more than what is enough
155.
I have cursed others
126.
I manage people well
156.
I forgive others when they have hurt
me by mistake
127.
I make people enthusiastic about
157.
I accept the strengths and weaknesses
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
158.
of others
185.
I express ideas that help people realize
new things
I always obey laws, even if I am
unlikely to get caught
186.
I refuse to talk to others I have argued
with
187.
I try to fool others
188.
I let people talk to me as a friend
159.
I am a role model for others
160.
I encourage people to think critically
161.
I give hope to others when they have
lost hope
189.
I am truthful in what I do
190.
I help people realize their potential
162.
I boost others' self-esteem
191.
163.
I talk about others in their absence
164.
I speak on behalf of others when they
cannot defend themselves
If a close person is ill and cannot go to
the doctor by him- or herself, I take him
or her to the doctor
192.
I wish people to achieve their goals
I try to make things better when I have
hurt someone
193.
I apologise if I have made a mistake
194.
I tell stories with a moral
166.
I help others with their work
195.
I pay my debts
167.
I am a person others love to be around
196.
I do things my way
168.
I admit when I do not know something
197.
I speak calmly
169.
I give advice to others about their
future
198.
I make people speak freely
170.
I love others even if we have
differences
199.
I make people feel special
200.
I am better than others
165.
171.
I oppose other people’s lifestyle
201.
I warn others about dangers
172.
I help others solve their problems
202.
I value the little things in life
173.
I hate it when others fight
203.
I make others feel stupid
174.
I give others the wrong impression of
myself
204.
It is hard for me to get rid of a
disturbing thought
175.
I attract the attention of people around
me in a positive way
205.
I share what I have with others
206.
176.
I understand the viewpoint of others
I have sometimes dropped litter on the
street
177.
I share helpful ideas
207.
I behave in an arrogant manner
178.
I like having everything my way
208.
I help others cope with their problems
179.
I refuse help from others
209.
I show the way to others
180.
I encourage people to develop
210.
181.
I encourage people to get along with
each other
I turn people away from the bad things
they do
211.
I treat all people with respect
212.
I take good care of my things
213.
I help people develop
214.
I make people believe in their own
abilities
215.
I consider others’ needs
182.
I like the things that can help people
live better
183.
I make others better persons
184.
I help others to make peace with each
other
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
216.
I take sides with people from my own
background
246.
I have prevented others from
succeeding in what they are doing
217.
I tell other people when I am grateful
247.
218.
I refuse accepting corrections from
others
I am patient with people with a different
way of living
248.
I have taken things that do not belong
to me
249.
I feel sympathy for people who have
problems
250.
I am full of love towards others
251.
I cheat
219.
I hate some people
220.
I guide people so they make the right
decisions
221.
I put others first to maintain a peaceful
relationship
222.
I make up false stories
252.
I am jealous of others with good fortune
223.
I focus on others' weak points
253.
224.
I let someone else be punished for my
wrongdoings
I listen patiently when others want to
share their feelings
254.
I find solutions to problems that others
have
I do as if I like people whom I actually
do not like
255.
I empower others by giving them
knowledge
256.
I work well with others
257.
I adapt to any situation
258.
I behave in an appropriate manner
259.
I treat all people equally
260.
I keep others' secrets
225.
226.
I earn other people's respect
227.
I accept people with their problems
228.
I always do as I say
229.
I am grateful for what life brings me
230.
I am able to deal with different
problems of different people
231.
I speak politely to others
261.
232.
I get along with others in my
community
I teach others about what is important
in life
262.
I make sure that others are well
233.
I have led others into troubles
263.
I only care about my own things
234.
I shout when I have an argument with
someone
264.
I dislike it when others get ahead in life
265.
I make time for others
235.
I give everyone a chance
266.
236.
I make an effort to have good
relationships with others
I have conversations with others from
which they can benefit
267.
237.
I make life difficult for others
I support others when they experience
problems
238.
I distrust other people’s opinions
268.
239.
I threaten people
I tell educative stories to younger
people
240.
I provide for those who are in need
269.
I provide help for the needs of my
community
241.
I am respected by many people
270.
I make fun of others
242.
I am honest with other people
271.
I behave according to my roles
243.
I am satisfied with what I have in life
272.
I insult people
244.
I want to have more than others
273.
245.
I motivate others to improve
I get sad when someone I care about is
sad
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
274.
I like seeing people progress in life
284.
I enjoy working with others
275.
I help others to succeed
285.
I scold people
276.
I use others for my own purposes
286.
I am someone others look up to
277.
There has been at least one occasion
when I failed to return something I
borrowed
287.
I love my close people even if they
have hurt me
288.
I ignore people
278.
I am a friend one can rely on
289.
I talk to others to resolve differences
279.
I am honest when giving my opinion
290.
I relate well to others
280.
I am irritated by people who ask
favours
281.
I comfort others when they are going
through hard times
282.
I am humble to the people I live with
283.
I tell others what to do without being
asked to do so
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
INSTRUCTIONS PART 2
This questionnaire contains a number of items relevant to your personality characteristics.
Please consider each item and decide to what extent it describes your personality
characteristics. Please indicate your disagreement or agreement by darkening the circle on your
answer sheet that best indicates your answer. There are no correct or wrong answers to these
questions. Please answer all the questions honestly. Thank you for your cooperation.
291.
Rules and laws should be strictly
enforced and should be without
exception.
304.
Parents should tolerate their children's
"rebellious" behaviour.
292.
When I see others feeling distressed, I
am easily affected too.
305.
People may think I am a person who
only follows conventions and
regulations
293.
My friends think that I am a trustworthy
person.
306.
I try my best to listen to my parents out
of filial piety.
294.
I try my best to maintain harmony in my
family because I believe that if a family
lives in harmony, all things will prosper.
307.
I frequently buy snacks.
308.
Students should concentrate on their
studies and not get distracted by what
is happening in the society.
309.
Blood is thicker than water, and no
matter what, one's feelings for one's
family are always stronger than for
outsiders.
310.
Children do not have to follow their
parents' wishes when choosing a
partner for marriage.
295.
If a friend or relative was hospitalized, I
would definitely go visit him/her.
296.
When I accomplish something
important, I try hard not to get too
excited, because I know that success
does not happen very often.
297.
At times, my friends remark that I am
too self-centred.
298.
Other people say that I am
understanding.
311.
When I talk to people, I seldom notice
whether I am offending them.
299.
I like to store old things for future use.
312.
300.
In order to avoid offending others, it is
best not to show off too much.
A woman's chastity is more important
than her life.
313.
Ancestral sacrifices, weddings,
funerals, etc. should be conducted in
keeping with their traditional forms and
etiquette, and should not be modified
casually.
314.
I will listen to other people's opinions
before making my decisions.
301.
To maintain a pure and simple culture,
eccentric clothes and hairstyles should
be strictly banned.
302.
I always discuss with others before I
reject any particular suggestion.
303.
I am good at sensing other people's
emotions.
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong & Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
TURN TO NEXT PAGE
315.
I have followed rules carefully since I
was young in order to avoid
punishment.
330.
Education is a sacred profession;
therefore teachers should not mind too
much about their pay.
316.
After I have been treated to a meal, I
will try to return the favour as soon as
possible.
331.
When I am interacting with others, I
seldom notice whether I am giving
them a hard time.
317.
It is unworthy for me to talk to or
socialize with people whose
educational level or social status are
lower than mine.
332.
I follow the saying that "Those who are
contented are always happy" as a
principle in life.
333.
318.
I cannot stand people who spend
money like water.
When facing a dilemma, I can always
arrive at a compromise.
334.
319.
It is not difficult for me to say "No" to
my friends' requests.
I found it difficult to put myself in others'
shoes to consider their points of view.
335.
320.
When I shop, I usually consider my
own likings only, and am seldom
influenced by others.
I never hesitate or feel sorry when I
decide to throw away old things.
336.
No matter who supports the family
financially, the man is still the head of
the household.
337.
During the holiday seasons, I always
visit relatives and friends for bonding or
strengthening our relationships.
338.
I easily get into conflict with other
people.
339.
If a dispute cannot be resolved, a
family elder should be invited to act as
an arbiter to uphold justice.
340.
I usually can maintain peace of mind.
341.
I believe most famous people have real
substance and seldom rely on others to
blow their trumpets.
342.
I always try hard to get along well with
others.
343.
It is acceptable for subordinates to
voice their opinions if their superiors
made a mistake.
344.
If a friend or a relative was
hospitalized, I might not necessarily go
visit him/her.
It would be great if everyone had a
similar way of thinking or a similar
value system.
345.
In order to save money, I take public
transport whenever I go out.
There is no barrier between me and
other people.
346.
My mind is at peace, and I have few
desires.
321.
Wherever I am, the atmosphere is
always lively.
322.
I can remember how I have spent my
money even if it were only a few
dollars.
323.
Kids that deserve the most praise are
those who obey the rules just as adults
do.
324.
I often wish everyone would talk with
me in a straightforward and
unambiguous way.
325.
One can avoid making serious
mistakes by always following tradition.
326.
To avoid mistakes in life, the best thing
to do is to listen to the elders'
suggestions.
327.
When a friend borrows something from
me and does not return it, I often feel
uneasy about asking him/her to give it
back.
328.
329.
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong & Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
TURN TO NEXT PAGE
always plan and organize all the
details.
347.
I actively try to understand other
people's needs.
348.
Even though I am not very rich, I would
still prefer to buy the most expensive
out of several brands of the same
product.
349.
I will keep new information to myself
and do not feel the need to share it with
others.
357.
I am very demanding on myself; it
would be great if everyone else was
like that.
358.
I am very thrifty even when I am using
public property.
359.
When dealing with organizations,
things can work out more smoothly
through the connections of friends
working inside.
360.
When people show me respect, I
should show them more respect in
return.
350.
When I see strangers getting hurt, I feel
hurt too.
351.
Even when I have new clothes, I
continue to wear something old and
save the new ones for important
occasions.
361.
352.
I like to save money for future
necessities.
I do not insist on making detailed plans
and arrangements for my work.
362.
Cohabitation is acceptable.
353.
It is a virtue to tolerate everything.
364.
354.
I am open and sincere when relating to
people.
I believe traditional ideas or concepts
should not be torn down.
364.
There is no stigma about marrying a
divorced person.
355.
I have a much stricter sense of right or
wrong than most people.
356.
No matter what kind of work I take up, I
END – THANK YOU
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong & Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
Psychology
School of Human & Community Development
University of the Witwatersrand
Private Bag 3, WITS, 2050
Tel: (011) 717 4500
Fax: (011) 717 4559
Dear Student
Hi! We are currently conducting research on the interpersonal relatedness factors of personality. This
study forms part of a larger study on personality in South Africa. Part of this research requires your
responses on the attached questionnaire. It should take you approximately 1½ hours to complete the
questionnaire. Please ensure that you answer on the answer sheet. I understand that this is a
substantial investment of your time. However your response is valuable as it will contribute towards a
South African understanding of personality and will have an impact on research nationally and
internationally. I would therefore like to invite you to participate in this research.
Your responses will remain confidential and anonymity is guaranteed. At no time will I know who you are
since the questionnaire requires no identifying information. Completion and return of the questionnaire will
be considered to indicate permission for me to use your responses for the research project. Should you
choose not to participate, this will not be held against you in any way. As I am only interested in group
trends, and have no way of linking any individual’s identity to a particular questionnaire, I will not be able
to give you individual feedback. Feedback in the form of a one-page summary sheet will be available on
request approximately six months after completion of this questionnaire. If you have any further questions
or require feedback on the progress of the research, please feel free to contact me. My contact details
appear below my signature. The data from this study will be used in conjunction with data collected from
other South African samples to determine the value of the South African Personality Inventory.
Thank you for considering taking part in the research project. Please detach and keep this sheet.
__________________
Dr. S. Laher
011 717 4532 / [email protected]
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong&Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
TURN TO NEXT PAGE
SH2
DIRECTIONS
This booklet contains statements about your interests, preferences, everyday thoughts and
behaviours and how you feel about things. You need to respond to all statements in the
booklet. The questionnaire contains 364 statements. Please use an HB pencil to complete the
questionnaire. Darken the circle that corresponds to the number that best indicates your
answer:
1. Strongly Disagree (1)
2. Disagree (2)
3. Somewhat Disagree / Somewhat Agree (3)
4. Agree (4)
5. Strongly Agree (5)
When the administrator asks you to start, answer each statement. Please keep the following
things in mind:
1. Read each statement and choose the answer that best describes you. There are no “right”
or “wrong” answers; just mark what is true for you.
2. If you want to change an answer, please erase the original one completely and darken the
circle of the number that best indicates your answer.
3. Please answer truthfully.
4. Complete the biographical information.
Please ask now if you have any questions.
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong&Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
TURN TO NEXT PAGE
SH2
INSTRUCTIONS PART 1
This questionnaire contains a number of items relevant to your personality characteristics.
Please consider each item and decide to what extent it describes your personality
characteristics. Please indicate your disagreement or agreement by darkening the circle on your
answer sheet that best indicates your answer. There are no correct or wrong answers to these
questions. Please answer all the questions honestly. Thank you for your cooperation.
1.
Rules and laws should be strictly
enforced and should be without
exception.
10.
In order to avoid offending others, it is
best not to show off too much.
2.
When I see others feeling distressed, I
am easily affected too.
11.
To maintain a pure and simple culture,
eccentric clothes and hairstyles should
be strictly banned.
3.
My friends think that I am a trustworthy
person.
12.
I always discuss with others before I
reject any particular suggestion.
4.
I try my best to maintain harmony in my
family because I believe that if a family
lives in harmony, all things will prosper.
13.
I am good at sensing other people's
emotions.
14.
5.
If a friend or relative was hospitalized, I
would definitely go visit him/her.
Parents should tolerate their children's
"rebellious" behaviour.
15.
6.
When I accomplish something
important, I try hard not to get too
excited, because I know that success
does not happen very often.
People may think I am a person who
only follows conventions and
regulations
16.
I try my best to listen to my parents out
of filial piety.
17.
I frequently buy snacks.
18.
Students should concentrate on their
studies and not get distracted by what
is happening in the society.
19.
Blood is thicker than water, and no
7.
At times, my friends remark that I am
too self-centred.
8.
Other people say that I am
understanding.
9.
I like to store old things for future use.
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong&Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
TURN TO NEXT PAGE
matter what, one's feelings for one's
family are always stronger than for
outsiders.
unambiguous way.
35.
One can avoid making serious
mistakes by always following tradition.
36.
To avoid mistakes in life, the best thing
to do is to listen to the elders'
suggestions.
37.
When a friend borrows something from
me and does not return it, I often feel
uneasy about asking him/her to give it
back.
38.
If a friend or a relative was
hospitalized, I might not necessarily go
visit him/her.
39.
In order to save money, I take public
transport whenever I go out.
40.
Education is a sacred profession;
therefore teachers should not mind too
much about their pay.
41.
When I am interacting with others, I
seldom notice whether I am giving
them a hard time.
20.
Children do not have to follow their
parents' wishes when choosing a
partner for marriage.
21.
When I talk to people, I seldom notice
whether I am offending them.
22.
A woman's chastity is more important
than her life.
23.
Ancestral sacrifices, weddings,
funerals, etc. should be conducted in
keeping with their traditional forms and
etiquette, and should not be modified
casually.
24.
I will listen to other people's opinions
before making my decisions.
25.
I have followed rules carefully since I
was young in order to avoid
punishment.
26.
After I have been treated to a meal, I
will try to return the favour as soon as
possible.
42.
27.
It is unworthy for me to talk to or
socialize with people whose
educational level or social status are
lower than mine.
I follow the saying that "Those who are
contented are always happy" as a
principle in life.
43.
28.
I cannot stand people who spend
money like water.
When facing a dilemma, I can always
arrive at a compromise.
44.
I found it difficult to put myself in others'
shoes to consider their points of view.
29.
It is not difficult for me to say "No" to
my friends' requests.
45.
I never hesitate or feel sorry when I
decide to throw away old things.
30.
When I shop, I usually consider my
own likings only, and am seldom
influenced by others.
46.
No matter who supports the family
financially, the man is still the head of
the household.
31.
Wherever I am, the atmosphere is
always lively.
47.
32.
I can remember how I have spent my
money even if it were only a few
dollars.
During the holiday seasons, I always
visit relatives and friends for bonding or
strengthening our relationships.
48.
I easily get into conflict with other
people.
33.
Kids that deserve the most praise are
those who obey the rules just as adults
do.
49.
If a dispute cannot be resolved, a
family elder should be invited to act as
an arbiter to uphold justice.
34.
I often wish everyone would talk with
me in a straightforward and
50.
I usually can maintain peace of mind.
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong&Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
TURN TO NEXT PAGE
51.
I believe most famous people have real
substance and seldom rely on others to
blow their trumpets.
52.
I always try hard to get along well with
others.
53.
It is acceptable for subordinates to
voice their opinions if their superiors
made a mistake.
54.
It would be great if everyone had a
similar way of thinking or a similar
value system.
55.
There is no barrier between me and
other people.
56.
My mind is at peace, and I have few
desires.
57.
I actively try to understand other
people's needs.
58.
59.
Even though I am not very rich, I would
still prefer to buy the most expensive
out of several brands of the same
product.
I will keep new information to myself
and do not feel the need to share it with
others.
60.
When I see strangers getting hurt, I feel
hurt too.
61.
Even when I have new clothes, I
continue to wear something old and
save the new ones for important
occasions.
62.
I like to save money for future
necessities.
63.
It is a virtue to tolerate everything.
64.
I am open and sincere when relating to
people.
65.
I have a much stricter sense of right or
wrong than most people.
66.
No matter what kind of work I take up, I
always plan and organize all the
details.
67.
I am very demanding on myself; it
would be great if everyone else was
like that.
68.
I am very thrifty even when I am using
public property.
69.
When dealing with organizations,
things can work out more smoothly
through the connections of friends
working inside.
70.
When people show me respect, I
should show them more respect in
return.
71.
I do not insist on making detailed plans
and arrangements for my work.
72.
Cohabitation is acceptable.
73.
I believe traditional ideas or concepts
should not be torn down.
74.
There is no stigma about marrying a
divorced person.
This questionnaire is adapted from Cross-Cultural (Chinese) Personality Assessment Inventory-2 (CPAI-2)
©2001 The Chinese University of Hong Kong&Institute of Psychology, Academia Sinica, China
Fanny M. Cheung, Kwok Leung, Song Weizheng, and Zhang Jianxin (The Author)
All rights reserved. No parts of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval
system, without the prior written permission of The Author.
TURN TO NEXT PAGE
INSTRUCTIONS PART 2
Please read each of the statements below and indicate your disagreement or agreement by
darkening the circle on your answer sheet that best indicates your answer:
75.
I help people grow
76.
I am loyal to others
77.
I greet others
78.
I accept change
79.
when they are away
101.
I maintain relations with others for a
long time
102.
I think about my options before I make
a choice
I boast about things that make me
better than others
103.
I play a leading role in the lives of
people close to me
80.
I accept others
104.
I have a low opinion of others
81.
I am suspicious of what people close to
me do
105.
I give advice that people would keep
for a long time
82.
I tell the truth
106.
I hurt others
83.
I help people find their way in life
107.
I teach people ways of doing things
84.
I listen to other people's problems
108.
I cause arguments between others
85.
I am a nice person
109.
I remain serious when jokes are told
86.
I challenge people in front of others
110.
I keep my things for myself
87.
I support others when they have to
make difficult decisions
111.
I forgive easily
88.
When I hear people talking privately, I
avoid listening
112.
I wish others to be successful
113.
I uplift others
114.
I behave differently from how I really
feel
115.
I care for others like for my own family
116.
I have disliked it when others are
happy
117.
I regard my background as being better
than the backgrounds of others
118.
I have some bad habits
89.
I spread rumours about others
90.
I hide from others who I really am
91.
I criticise others' mistakes
92.
I sometimes regret my decisions
93.
I protect others
94.
I fake a smile when I have to
95.
I take revenge on people who have
hurt me
119.
96.
I want to sort things out in a peaceful
manner
I have done things that I do not tell
other people about
120.
97.
I provide guidance to help others
succeed
I pass my passion about things on to
others
121.
I treat weaker people badly
98.
I give others emotional support
122.
I take others’ feelings into account
99.
I interfere in the lives of others
123.
I talk to people in a respectful way
100.
I look after other people’s belongings
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
124.
I make others feel comfortable
151.
I only think of myself
125.
Once I have made up my mind, other
people cannot change my opinion
152.
I am open to people with different
opinions
126.
I show people how to live their life in a
good way
153.
I abuse my power over others
154.
I have humiliated others
127.
I can use many different ways of doing
things
155.
I provide directions to others about
what is right and wrong
128.
I give people directions in life
156.
I like doing things for others
129.
I criticise people when I do not
understand them
157.
I live by my principles
130.
I continue with my work if I am
motivated
158.
I have doubts about my ability to
succeed in life
131.
I make people do things for me
159.
I support people close to me in
whatever they choose to do
132.
I am just an average person
160.
133.
I have done things that are against the
law
I lead people by showing good
behaviour
161.
134.
I am happy when I hear bad news
about people whom I do not like
I keep people's attention when I explain
something
162.
I accept apologies
135.
I tell people how they can cope with
difficulties
163.
I am very confident of my judgments
136.
I start fights with others
164.
I respect others’ opinions
137.
I advise others when they are in
unfamiliar situations
165.
I care about other people's future
166.
I provoke others
138.
I get involved in other's affairs
167.
I am a good leader
139.
I accept the advice of others
168.
I have been violent in the past
140.
I guide people in life
169.
I resolve conflicts between people
141.
I take advantage of weaker people
170.
I help others when they are in need
142.
I have hurt others with my words
171.
143.
I sometimes tell lies if I have to
I believe that many people have bad
intentions
144.
I believe that there is nothing to learn
from other people
172.
I am easy to talk to
173.
I can see others’ point of view
174.
I create tension between others
175.
I am sensitive to other people’s feelings
176.
I disappoint others
177.
My first impressions of people usually
turn out to be right
178.
I hug people when I am happy about
them
179.
I want people to live in peace
180.
I look down on others
145.
I am approachable when others need
help
146.
I am a person everybody listens to
147.
I discriminate against people
148.
I have taken things from others using
force
149.
It is hard for me to break my bad habits
150.
I say bad things about others while
they are not there
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
181.
I indulge others
212.
I make others feel at home
182.
I value others for what they are
213.
183.
I do what is expected of me
I have so much pride that I find it
difficult to accept others
184.
I think I am more important than others
214.
I gossip
185.
I have embarrassed people in front of
others
215.
I help people to solve their arguments
216.
I find ways to help people even if it is
difficult
217.
I influence people in important ways in
their lives
218.
I am a true friend to others
219.
I ask people if they are all right
220.
I encourage others when they face
difficulties
186.
I am careful about my way of dressing
187.
I show humanity towards others
188.
I have a clean mind about my actions
189.
I treat others in a careful way
190.
I only see the bad sides of people
191.
I favour some people above others
192.
I find it difficult to confide in others
221.
I am friendly towards others
193.
I treat some people badly because of
their background
222.
I stand by what I say
223.
194.
I am an example for others
I show people the right way when they
are unsure about something
195.
I give things to people without
expecting anything in return
224.
I discuss people's personal affairs with
others
196.
I do the right thing
225.
I look after the safety of others
197.
I listen carefully to understand what
others are saying
226.
I am a source of inspiration to people
227.
I come home at the expected time
198.
I take responsibility for my mistakes
228.
I give my attention to others
199.
I want more than what is enough
229.
I have cursed others
200.
I manage people well
230.
201.
I make people enthusiastic about
different subjects
I forgive others when they have hurt
me by mistake
231.
202.
I am kind to others
I accept the strengths and weaknesses
of others
203.
I keep my promises
232.
204.
I avoid arguments
I express ideas that help people realize
new things
205.
I make people feel involved when I
explain something
233.
I am a role model for others
234.
I encourage people to think critically
206.
I make jokes about other people
235.
207.
I make people feel weak when they are
around me
I give hope to others when they have
lost hope
236.
I boost others' self-esteem
208.
I can share in someone's emotions
237.
I talk about others in their absence
209.
I choose the people I want to speak to
238.
210.
I allow others to ask me questions
I speak on behalf of others when they
cannot defend themselves
211.
I have frightened others
239.
I try to make things better when I have
hurt someone
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
240.
I help others with their work
268.
I tell stories with a moral
241.
I am a person others love to be around
269.
I pay my debts
242.
I admit when I do not know something
270.
I do things my way
243.
I give advice to others about their
future
271.
I speak calmly
272.
I make people speak freely
244.
I love others even if we have
differences
273.
I make people feel special
245.
I oppose other people’s lifestyle
274.
I am better than others
246.
I help others solve their problems
275.
I warn others about dangers
247.
I hate it when others fight
276.
I value the little things in life
248.
I give others the wrong impression of
myself
277.
I make others feel stupid
278.
I attract the attention of people around
me in a positive way
It is hard for me to get rid of a
disturbing thought
279.
I share what I have with others
250.
I understand the viewpoint of others
280.
251.
I share helpful ideas
I have sometimes dropped litter on the
street
252.
I like having everything my way
281.
I behave in an arrogant manner
253.
I refuse help from others
282.
I help others cope with their problems
254.
I encourage people to develop
283.
I show the way to others
255.
I encourage people to get along with
each other
284.
I turn people away from the bad things
they do
256.
I like the things that can help people
live better
285.
I treat all people with respect
286.
I take good care of my things
249.
257.
I make others better persons
287.
I help people develop
258.
I help others to make peace with each
other
288.
I make people believe in their own
abilities
259.
I always obey laws, even if I am
unlikely to get caught
289.
I consider others’ needs
290.
260.
I refuse to talk to others I have argued
with
I take sides with people from my own
background
291.
I tell other people when I am grateful
261.
I try to fool others
292.
262.
I let people talk to me as a friend
I refuse accepting corrections from
others
263.
I am truthful in what I do
293.
I hate some people
264.
I help people realize their potential
294.
265.
If a close person is ill and cannot go to
the doctor by him- or herself, I take him
or her to the doctor
I guide people so they make the right
decisions
295.
I put others first to maintain a peaceful
relationship
266.
I wish people to achieve their goals
296.
I make up false stories
267.
I apologise if I have made a mistake
297.
I focus on others' weak points
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
298.
I let someone else be punished for my
wrongdoings
323.
I feel sympathy for people who have
problems
299.
I find solutions to problems that others
have
324.
I am full of love towards others
325.
I cheat
300.
I earn other people's respect
326.
I am jealous of others with good fortune
301.
I accept people with their problems
327.
302.
I always do as I say
I listen patiently when others want to
share their feelings
303.
I am grateful for what life brings me
328.
304.
I am able to deal with different
problems of different people
I do as if I like people whom I actually
do not like
329.
I empower others by giving them
knowledge
330.
I work well with others
331.
I adapt to any situation
305.
I speak politely to others
306.
I get along with others in my
community
307.
I have led others into troubles
332.
I behave in an appropriate manner
308.
I shout when I have an argument with
someone
333.
I treat all people equally
334.
I keep others' secrets
309.
I give everyone a chance
335.
310.
I make an effort to have good
relationships with others
I teach others about what is important
in life
336.
I make sure that others are well
311.
I make life difficult for others
337.
I only care about my own things
312.
I distrust other people’s opinions
338.
I dislike it when others get ahead in life
313.
I threaten people
339.
I make time for others
314.
I provide for those who are in need
340.
315.
I am respected by many people
I have conversations with others from
which they can benefit
316.
I am honest with other people
341.
317.
I am satisfied with what I have in life
I support others when they experience
problems
318.
I want to have more than others
342.
I tell educative stories to younger
people
319.
I motivate others to improve
343.
320.
I have prevented others from
succeeding in what they are doing
I provide help for the needs of my
community
344.
I make fun of others
345.
I behave according to my roles
346.
I insult people
351.
There has been at least one occasion
when I failed to return something I
borrowed
352.
I am a friend one can rely on
353.
I am honest when giving my opinion
321.
I am patient with people with a different
way of living
322.
I have taken things that do not belong
to me
347.
I get sad when someone I care about is
sad
348.
I like seeing people progress in life
349.
I help others to succeed
350.
I use others for my own purposes
(C) 2010 all rights reserved SAPI Project Team
TURN TO NEXT PAGE
354.
I am irritated by people who ask
favours
355.
I comfort others when they are going
through hard times
356.
I am humble to the people I live with
357.
I tell others what to do without being
asked to do so
358.
I enjoy working with others
359.
I scold people
360.
I am someone others look up to
361.
I love my close people even if they
have hurt me
362.
I ignore people
364.
I talk to others to resolve differences
364.
I relate well to others
END – THANK YOU
(C) 2010 all rights reserved SAPI Project Team
197
Reference List
Abrahams, F. (1996). The Cross-Cultural Comparability of the 16 Personality
FactorInventory (16PF). Unpublished Doctoral Thesis. South Africa: UNISA.
Abrahams, F. &Mauer, K. F. (1999). Qualitative and statistical impacts of home language on
responses to the items of the Sixteen Personality Factor Questionnaire (16PF) in South
Africa. South African Journal of Psychology 29, 76-86.
Allik, J., & McCrae, R. R. (2004). Toward a geography of personality traits. Patterns of
profile across 36 cultures. Journal of Cross-Cultural Psychology, 35, 13-28.
Allport, G.W. (1937). Personality: A psychological interpretation. New York: Holt.
Allport, G.W. (1961). Pattern and growth in personality. Oxford, England: Holt, Reinhart &
Winston.
Allport, G. W. & Odbert, H. S. (1936). Trait names: A psycholexical study. Psychological
Monographs,47, 211.
Anastasi, A. &Urbina, S. (2007). Psychological Testing. New Jersey: Prentice Hall.
Ashton, M. C. & Lee, K. (2005). The lexical approach to the study of personality structure:
Toward the identification of cross-culturally replicable dimensions of personality
variation. Journal of Personality Disorders, 19, 303-308.
Ashton, M. C., Lee, K., Perugini, M., Szarota, P., De Vries., R. E., Di Blas, L., Boies, K., &
De Raad, B. (2004). A six-factor structure of personality-descriptive adjectives:
solutionsfrom psycholexical studies in seven languages. Journal of Personality and
SocialPsychology, 86, 356 – 366.
Babbie, E. and Mouton, J. (2004). The Practice of Social Research. Cape Town: Oxford
University Press.
Barbour, R. S. (2000). The role of qualitative research in broadening the ‘evidence base’ for
clinical practice. Journal of Evaluation in Clinical Practice, 6, 155-163.
198
Bedell, B., Van Eeden, R. & Van Staden, F. (1999). Culture as a moderator variable in
psychological test performance: Issues and trends in South Africa. South African
Journal of Industrial Psychology, 25, 1 -7.
Berry, J. W. & Sam, D. (1997). Acculturation and adaptation. In J. W. Berry, M. H. Seagall&
C. Kagitsibasi (Eds.), Handbook of cross-cultural psychology (Vol. 3). Boston: Allyn&
Bacon.
Boyatzis, R. E. (1998). Transforming qualitative information: Thematic analysis and code
development. London: SAGE Publications.
Branco e Silva, L. M. &Laher, S. (2010). Exploring the collective dimension of personality
using the SAPI and CPAI-2 in a sample of South African students.Ife PsychologIA, 20,
19-48.
Braun, V. & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research
in Psychology, 3, 77-101.
Carmines, E. G. & Zeller, R. A. (1979). Reliability and Validity Assessment. London: SAGE
Publications.
Cattell, R. B. (1957). Personality and motivation: Structure and measurement. Journal of
Personality Disorders, 19, 53-67.
Cheung, F.M. (2004). Use of Western and indigenously developed personality tests in Asia.
Applied Psychology: An International Review, 53, 173-191.
Cheung, F.M. (2006). CPAI-2 Personality Scale Description. Available from Fanny Cheung,
Department of Psychology, Chinese University of Hong Kong.
Cheung, F., Cheung, S., Leung, K., Ward, C., & Leong, F. (2003). The English version of the
Chinese Personality Assessment Inventory. Journal of Cross-Cultural Psychology, 34,
433-452.
Cheung, F. M., Cheung, S. F., Zhang, J., Leung, K., Leong, F. & Huiyeh, K. (2008).
Relevance of Openness as a personality dimension in Chinese culture: Aspects of its
cultural relevance. Journal of Cross-Cultural Psychology, 39, 81-108.
199
Cheung, F.M. & Leung, K. (1998). Indigenous personality measures: Chinese examples.
Journal of Cross-Cultural Psychology, 29, 32-62.
Cheung, F.M., Leung, K., Fan, R., Song, W.Z., Zhang, J.X. & Zhang, J.P. (1996).
Development of the Chinese Personality Assessment Inventory (CPAI). Journal of
Cross-Cultural Psychology, 27, 181-199.
Cheung, F.M., Leung, K., Zhang, J.X., Sun, H.F., Gan, Y.Q., Song, W.Z. &Xie, D. (2001).
Indigenous Chinese personality constructs: Is the Five-Factor Model complete? Journal
of Cross-Cultural Psychology, 32, 407-433.
Cheung, S. F., Cheung, F M., Howard, R. & Lim, Y-H. (2006). Personality across the ethnic
divide in Singapore: Are “Chinese traits” uniquely Chinese? Personality and Individual
Differences, 41, 467-477.
Church, T.A. (2000). Culture and personality: Toward an integrated cultural trait psychology.
Journal of Personality, 68, 651-704.
Church, A.T. &Lonner, W.J. (1998). The cross-cultural perspective in the study of
personality. Journal of Cross-Cultural Psychology, 29, 32-62.
Cohen, R. J. &Swerdlik, M. E. (2005). Psychological Testing and Assessment (6thed).
Boston: McGraw-Hill.
Costa, P.T. & McCrae, R.R. (1992). NEO-PI-R Professional Manual. Florida, Psychological
Assessment Resources, Inc.
Costa, P.T. & McCrae, R.R. (2008). The Revised NEO Personality Inventory. In G.J. Boyle,
G. Matthews & D.H. Saklofske (Eds.), The Sage Handbook of Personality Theory and
Assessment. London: Sage Publications.
Costa, P. T., Terracciano, A. & McCrae, R. R. (2001). Gender differences in personality traits
across cultures: Robust and surprising findings. Journal of Personality and Social
Psychology, 81, 322-331.
Creswell, J. W. (2009). Research design: Qualitative, Quantitative, and Mixed Methods
Approaches (3rd Ed.). Thousand Oakes, CA: SAGE Publications.
200
Cross, S. E., & Markus, H. R. (1999). The cultural constitution of personality. In L. Pervin,
O. John (Eds.), Handbook of personality theory and research (2nd Ed.). New York:
Guilford Press.
Croxford, S.A. (2009).Exploring gender differences across personality traits in a sample of
South African university students. Unpublished Honours dissertation (Psychology).
University of the Witwatersrand, Johannesburg.
Dalton, M. & Wilson, M. (2000) The relationship of the Five-Factor model of personality to
job performance for a group of middle-eastern expatriate managers. Journal of CrossCultural Psychology, 31, 250-258.
De Bruin, K., De Bruin, G. P., Dercksen, S. &Cilliers-Hartslief, M. (2005). Predictive
validity of general intelligence and Big Five measures for adult basic education and
training outcomes. South African Journal of Psychology, 35, 46-57.
Digman, J. M. &Takemoto-Chock, N. M. (1981).Factors In The Natural Language Of
Personality: Re-Analysis, Comparison, And Interpretation Of Six Major Studies.
Multivariate Behavioral Research, 16, 149-170.
Eaton, L. &Louw, J. (2002). Culture and self in SA: Individualism and collectivism
predictions. Journal of Social Psychology, 140, 210-217.
Ellemers, N., Spears, R. &Doosje, B. (2002). Self and social identity. Annual Review of
Psychology, 53, 161-186.
Ellingson, J. E., Sackett, P. R. & Hough, L. M. (1999). Social Desirability Corrections in
Personality Measurement: Issues of Applicant Comparison and Construct Validity.
Journal of Applied Psychology, 84, 155-166.
England, J. (1991). Personality assessment: A review with special reference to the South
African situation. Pretoria: Human Sciences Research Council.
Fan, Y. (2000). A classification of Chinese culture.Cross Cultural Management: An
International Journal, 7, 3-10.
Feingold, A. (1994). Gender differences in personality: A meta-analysis. Psychological
Bulletin, 116, 429-456.
201
Fine, M. (2002) Disruptive voices: the possibilities for feminist research. University of
MichiganPress.
Foxcroft, C. (1997). Psychological testing in South Africa: Perspectives regarding ethical and
fair practices. European Journal of Psychological Assessment, 13, 229-235
Foxcroft, C. (2004). Planning a psychological test in the multicultural South African context.
South African Journal of Industrial Psychology, 30, 8-15.
Foxcroft, C., &Roodt, G. (2009). An introduction to psychological assessment in the
SouthAfrican context (3rd Ed.). Cape Town: Oxford University Press, Southern Africa.
Franklin, D.R. (2009). Exploring Language Bias in the NEO-PI-R. Unpublished Masters
dissertation (Research Psychology). University of the Witwatersrand, Johannesburg.
Freeman, F. S. (1962). Theory and Practice of Psychological Testing. London: Holt, Rinehart
and Winston.
Friedman, H.S. &Schustack, M.W. (2006). Personality: Classic Theories and Modern
Research (3rd Ed.). USA: Pearson.
Goldberg, L.R. (1990). An alternative description of personality: The Big-Five factor
structure. Journal of Personality and Social Psychology, 59, 1216-1229.
Goldberg, L. R. (1995). What the hell took so long? Donald. W. Fiske and the Big Five factor
structure. In P. E. Shrout& S. T. Fiske (Eds.), Personality, research, methods and
theory: A Festschrift honoring Donald W. Fiske (pp. 29-43). Hillsdale, NJ: Lawrence
Erlbaum Associates.
Green, E.G.T., Deschamps, J-L. &Paez, D. (2005). Variation of Individualism and
Collectivism within and between 20 countries: A typological analysis. Journal of
Cross-Cultural Psychology, 36, 321-339.
Guion, R. M. &Highhouse, S. (2006). Essentials of Personnel Assessment and Selection.
New Jersey: Lawrence Erlbaum Associates.
Heaven, P. C. L., Conners, J. & Stones, C. R. (1994). Three or five personality dimensions?
An analysis of natural language terms in two cultures. Journal of Personality and
Individual Differences, 17, 181-189.
202
Heaven, P.C. & Pretorius, A. (1998). Personality structure among black and white South
Africans. The Journal of Social Psychology, 138, 664-666.ssment Resources, Inc.
Heuchert, J.W.P., Parker, W.D., &Stumpf, H.&Myburgh, C. P. H. (2000). The Five Factor
model of personality in South African college students. American Behavioural Scientist,
44, 112-125.
Ho, D. Y. F., Peng, S., Lai A. C., & Chan, S.F. (2001). Indigenization and beyond:
Methodological relationalism in the study of personality across cultural traditions.
Journal of Personality, 69, 925-953.
Holden, R. R. (2008). Underestimating the effects of faking on the validity of self-report
personality scales. Personality and Individual Differences, 44, 311–321.
Huck, S.W. (2009). Reading Statistics and Research (5th Ed.). Boston: Pearson.
Hyde, J. S. (2005). The gender similarities hypothesis. American Psychologist, 60, 581-592.
John, O.P., Robins, R.W. &Pervin, L.A. (2008). Handbook of personality: Theory and
research (3rd Ed.). New York: The Guilford Press.
Kline, P. (1993). Personality: The Psychometric View. New York: Routledge.
Kline, P. (1994). An easy guide to factor analysis. London: Routledge.
Laher, S. (2008). Structural equivalence and the NEO-PI-R: Implications for the applicability
of the Five Factor Model of personality in an African context. South African Journal of
Industrial Psychology, 34, 76-80.
Laher, S. (2010). The applicability of the NEO-PI-R and the CPAI-2 in the South African
context. Unpublished doctoral dissertation, University of the Witwatersrand,
Johannesburg.
Lalwani, A. K., Shavitt S. & Johnson, T. (2006). What is the relation between cultural
orientation and socially desirable responding? Journal of Personality and Social
Psychology, 90, 165-178.
Lanyon, R. I. & Goodstein, L. D. (1997). Personality Assessment (3rd Ed.). New York: John
Wiley & Sons.
203
Larsen, R.J. & Buss, D.M. (2008). Personality Psychology: Domains of Knowledge About
Human Nature (3rd Ed.). New York: McGraw-Hill International.
Leite, W. L. &Beretvas, N. S. (2005). Validation of Scores on the Marlowe-Crowne Social
Desirability Scale and the Balanced Inventory of Desirable Responding. Educational
and Psychological Measurement, 65, 140-154.
Li, A. & Bagger, J. (2006). Using the BIDR to Distinguish the Effects of Impression
Management and Self-Deception on the Criterion Validity of Personality Measures: A
Meta-Analysis. International Journal of Selection and Assessment, 14, 131-141.
Li, A. &Reb, J. (2009). A cross-nations, cross-cultures and cross-conditions analysis on the
equivalence of the Balanced Inventory of Desirable Responding. Journal of CrossCultural Psychology, 40, 214-233.
Lin, E. J., & Church, A.T. (2004). Are Indigenous Chinese Personality Dimensions CultureSpecific? An Investigation of the Chinese Personality Assessment Inventory in Chinese
American and European American Samples. Journal of Cross-Cultural Psychology, 35,
586-605.
Louw, D. J. (2001). Ubuntu and the challenges of multiculturalismin post-apartheid South
Africa. Retrieved January 27, 2012, from: http://www.phys.uu/nl ̴ unitwin/ubuntu.html.
Louw, D. J. (2002). Ubuntu: An African expression of the religious other. Polokwane:
University of the North.
Markus, H.R. &Kitayama, S. (1998). The cultural psychology of personality. Journal of
Cross-Cultural Psychology, 29, 63-87.
Matsimbi, W.E. (1997). Cross-cultural generalisability of the Five Factor model: a study on
South African white collar males. Unpublished Masters dissertation. University of the
Witwatersrand, Johannesburg.
McCrae, R.R. (2004). Human nature and culture: A trait perspective. Journal of Research in
Personality, 38, 3-14.
McCrae, R.R. (2001). Trait psychology and culture: Exploring intercultural comparisons.
Journal of Personality, 69, 819-846.
204
McCrae, R. R. &Allik, J. (2002). Introduction. In R. R. McCrae,& J. Allik (Eds). The FiveFactor Model of personality across cultures(pp. 1-5) New York: Kluwer
Academic/Plenum.
McCrae, R.R. & Costa, P.T. (1995). Domains and facets: hierarchical personality assessment
using the Revised NEO Personality Inventory. Journal of Personality Assessment, 64,
25-50.
McCrae, R.R., Costa, P.T. (1996). Toward a new generation of personality theories:
Theoretical contexts for the Five-Factor Model. In J.S. Wiggins (Ed.), The Five-Factor
Model of Personality: Theoretical perspectives (pp. 51-87). New York: Guilford Press.
McCrae, R.R. & Costa, P.T. (1997). Personality trait structure as a human universal.
American Psychologist, 52, 509-516.
McCrae, R.R., Costa, P.T., Del Pilar, G.H., Rolland, J.P. & Parker, W.D. (1998). Crosscultural assessment of the Five-Factor model. Journal of Cross-Cultural Psychology,29,
171-188.
McCrae, R.R. & Costa, P.T. Jr. (2008a). Empirical and theoretical status of the Fiver Factor
Model of personality traits. In G.J. Boyle, G. Matthews & D.H. Saklofske (Eds.), The
SAGE handbook of personality theory and assessment (pp. 273-294). London: Sage
Publications.
McCrae, R. R. & Costa, P. T. Jr. (2008b). The Five Factor Theory of personality. In O. P.
John, R. W. Robins, & L. A. Pervin (Eds). Handbook of personality: Theory and
research (3rd Ed., pp. 157-180). New York: Guilford Press.
McCrae, R. R. &Terracciano, A. & 78 Members of the Personality Profiles of Cultures
Project. (2005). Universal features of personality traits from the observers perspective:
Data from 50 cultures. Journal of Personality and Social psychology, 88, 547-561.
McCrae, R. R., Terracciano, A., Realo, A. &Allik, J. (2007). Climate warmth and national
wealth: Some culture-level determinants of national character stereotypes. European
Journal of Personality (in press). Published online in Wiley InterScience.
www.interscience.wiley.com.
205
McCrae, R.R., Yik, M.S., Trapnell, P.D., Bond, M.H. &Paulhus, D. (1998). Interpreting
personality profiles across cultures: Bilingual, acculturation, and peer rating studies of
Chinese undergraduates. Journal of Personality and Social psychology, 74, 1041-1055.
Meiring, D. (2006). South African Personality Inventory Project. Retrieved January 20, 2011,
from: http;//www.meiringd.co.za/sapi.htm
Meiring, D. (July, 2010). The SAPI in South Africa. Paper presented at the Annual Congress
of the Society of Industrial/Organisational Psychology in South Africa, The Forum
Campus, Bryanston, South Africa.
Meiring, D., Van de Vijver, F., Rothmann, S., &Barrick, M. R. (2005). Construct, item and
method bias of cognitive and personality tests in South Africa. South African Journal of
Industrial Psychology, 31, 1-8.
Monnette, D. R., Sullivan, T. J. &DeJong, C. R. (2010).Applied Social Research: A Tool for
the Human Services. Belmont, CA: Brooks/Cole.
Mpofu, E. (2001). Exploring the self concept in African culture. Journal of Genetic
Psychology, 155, 341-354.
Murphy, K. R. (2003). Validity generalization: a critical review. New Jersey: Lawrence
Erlbaum Associates Inc.
Murphy, K.R. &Davidshofer, C.O. (2005). Psychological testing: Principles and
Applications. New Jersey: Prentice Hall.
Nel, J. A. (2008). Uncovering personality dimensions in eleven different language groups in
South Africa: An exploratory study. Unpublished doctoral dissertation, University of the
North-West, Potchefstroom.
Nolte-Schamm, C. (2006). The African anthropology as a resource for reconciliation:
Ubuntu/Botho as a reconciliatory paradigm in South Africa. Scriptura, 93,370-383.
Nunally, J. C. & Bernstein, I. H. (1994). Psychometric theory. New York: McGraw-Hill.
Office of the President. (1998). Employment Equity Act. (Act No. 55, 1998). Government
Gazette Vol. 400, No. 19370, 19 October 1998.
206
Ones, D. S, Reiss, A. D., &Viswesvaran, C. (1996). Role of social desirability in personality
testing for personnel selection: The Red Herring. Journal of Applied Psychology, 81,
660-679.
Paulhus, D. L. (2002). Socially desirable responding: The evolution of a construct. In H.
Braun, D. N. Jackson, & D. E. Wiley (Eds.), The role of constructs in psychological and
educational measurement (pp. 67-88). Hillsdale, NJ: Erlbaum.
Paunonen, S.V. & Ashton, M.C. (1998). The structured assessment of personality across
cultures. Journal of Cross-Cultural Psychology, 29, 150-170.
Paunonen, S.V., Zeidner, M., Engvik, H.A., Oosterveld, P. &Maliphant, R. (2000). The
nonverbal assessment of personality in five cultures. Journal of Cross-Cultural
Psychology, 31, 220-239.
Pett, M.A., Lackey, N.R. & Sullivan, J.J. (2003). Making sense of factor analysis: the use of
factor analysis for instrument development in health care research. London: Sage
Publications.
Piedmont, R. L., Bain, E., McCrae, R. R., & Costa, P. T. Jr. (2002). The Applicability of the
Five Factor model in a sub-Saharan culture: The NEO-PI-R in Shona. In R. R. McCrae,
& J. Allik (Eds). The Five-Factor Model of personality across cultures (pp. 155-174)
New York: Kluwer Academic.
Piedmont, R. L., McCrae, R. R., Riemann R., &Angleitner, A. (2000). On the invalidity of
validity scales: Evidence from self-reports and observer ratings in volunteer samples.
Journal of Personality and Social Psychology, 78, 582-593.
Pillow, W. (2003). Confession, catharsis, or cure? Rethinking the uses of reflexivity as
methodological power in qualitative research. International Journal of Qualitative
Studies in Education, 16, 175-196.
Rhee, E., Uleman, J. S., Lee, H. K., & Roman, R. J. (1995). Spontaneous self-descriptions
and ethnic identities in individualistic and collectivistic cultures. Journal of Personality
and Social Psychology, 69, 142-152.
Robbins, S.P., Judge, T.A., Odendaal, A. &Roodt, G. (2009). Organisational Behaviour:
Global and Southern African Perspectives. Cape Town: Pearson.
207
Rosenthal, R. &Rosnow, R.L. (1991). Essentials of behavioral research: Methods and data
analysis. New York: McGraw Hill.
Roodt, G. & La Grange, L. (2001). Personality and cognitive ability as predictors of the job
performance of insurance sales people. South African Journal of Industrial Psychology,
27, 35-43.
Rumrill, P. D. Jr., Cook, B. G. & Wiley, A. L. (2011). Research in Special Education:
Designs, Methods, and Applications. Springfield, Illinois: Charles C Thomas
Publishers.
Rust, J. &Golombok, S. (1999). Modern psychometrics: The science of psychological
assessment (2ndEd.). London and New York: Routledge.
Sanderson, C. A. (2009). Social Psychology. USA: John Wiley & Sons.
Saucier, G. & Goldberg, L. R. (2001). Lexical studies of indigenous personality factors:
Premises, products and prospects. Journal of Personality, 69, 847-879.
Schneider, B. & Smith, D. B. (2004). Personality and Organizations. New Jersey: Lawrence
Erlbaum Associates.
Schultz, D. & Schultz, S. E. (2008). Theories of personality (9th Ed.). Belmont, CA:
Wadsworth.
Sehlapelo, M. & Terre Blanche, M. (1996). Psychometric testing in South Africa: Views
from above and below. Psychology in Society, 21, 49-59.
Silzer, R. F., Silzer, R. & Dowell, B. E. (2009).Strategy-driven talent management: a
leadership imperative. San Francisco: John Wiley & Sons.
Singelis, T. M., Triandis, H. C., Bhawuk, D. P S. &Gelfand, M. J. (1995). Horizontal and
vertical dimensions of individualism and collectivism: A theoretical and measurement
refinement. Cross-Cultural Research, 29, 240-275.
Spector, P. E. (1992). Summated rating scale construction: An introduction. Newbury Park,
CA: SAGE Publications.
208
Statistics South Africa (2011). Census 2011. Retrieved December 13, 2011, from:
http://www.statsa.gov.za/publications/populationstats.asp.
Swanson, S. (2007). Ubuntu: An African contribution to (re) research for/with a humble
togetherness. Journal of Contemporary Issues in Education, 2, 53-67.
Swartz, E. &Davies, R. (1997) Ubuntu - the spirit of African transformation management - a
review. Leadership & Organization Development Journal, 18, 290–294.
Taylor, I. A. (2000). The construct comparability of the NEO-PI-R questionnaire for Black
and white employees. Unpublished doctoral dissertation. University of the Orange Free
State, South Africa.
Taylor, G. W. & Ussher, J. M. (2001). Making Sense of S&M: A Discourse Analytic
Account. Sexualities, 4, 293-314.
Teferi, T.B. (2004). The application of the NEO-PI-R in the Eritrean context. Unpublished
Masters dissertation (Research Psychology). University of the Witwatersrand,
Johannesburg.
Triandis, H.C. (1995). Individualism and Collectivism. Boulder, CO: Westview Press.
Triandis, H. C. (2001). Individualism-Collectivism and personality. Journal of Perosnality,
69, 907-924.
Triandis, H.C. (1989). The self and social behaviour in differing contexts. Psychological
Review, 96, 508-520.
Tupes, E. C., &Christal, R. E. (1961). Recurrent personality factors based on trait ratings.
(ASD-TR-61-97). LacklandAirforce Base, TX: Aeronautical Systems Division,
Personnel Laboratory.
Uziel, L. (2010). Rethinking social desirability scales: From impression management to
interpersonally oriented self-control. Perspectives on Psychological Science, 5, 243262.
Van der Merwe, R. P. (2002). Psychometric Testing and Human Resource Management.
South African Journal of Industrial Psychology, 28, 77-86.
209
Van de Vijver, F. & Leung, K. (1997). Methods and data analysis for cross cultural
research. Newbury Park, CA: SAGE Publications.
Van de Vijver, F. & Leung, K. (2001).Personality in Cultural Context: Methodological
Issues. Journal of Personality, 69, 1007-1031.
Van de Vijver, F. &Rothmann, S. (2004). Assessment in multicultural groups: The South
African case. South African Journal of Industrial Psychology, 30, 1-7.
Van de Vijver, F. &Tanzer, N.K. (1997). Bias and equivalence in cross-cultural assessment:
An overview. European Review of Applied Psychology, 47, 263-280.
Van Eeden, R. &Prinsloo, C.H. (1997). Using the South African version of the 16PF in a
multicultural context. South African Journal of Psychology, 27, 151-159.
Verardi, S., Dahourou, D., Ah-Kion, J., Bhowon, U., Tseung, C. N., Amoussou-Yeye, D.,
Adjahouisso, M., Bouatta, C., Cissé, D. D., Mbodji, M., Barry, O., Minga, D. M.,
Ondongo, F., Tsokini, D., Rigozzi, C., Meyer de Stadelhofen, F. &Rossier, J. (2010).
Psychometric Properties of the Marlowe-Crowne Social Desirability Scale in Eight
African Countries and Switzerland. Journal of Cross-Cultural Psychology, 41, 19-34.
Vernon, P. E. (1957). Personality Tests and Assessments. London: Methuen & Co. Ltd.
Walford, G., Tucker, E. &Viswanathan, M. (2010). The SAGE Handbook of Measurement.
Thousand Oaks, CA: SAGE Publications.
Wallis, T. &Birt, M. (2003). A comparison of native and non-native English-speaking
groups’ understanding of the vocabulary contained in the 16PF (SA92). South African
Journal of Psychology, 33, 182-190.
Weiner, I.B. & Greene, R.L. (2008). Handbook of Personality Assessment. New Jersey: John
Wiley & Sons, Inc.
Weiner, I. B. &Craighead, W. E. (2010). The CorsiniEncyclopedia of Psychology (4th Ed).
Hoboken, NJ: John Wiley & Sons.
Weiten, W. (2007). Psychology: Themes and Variations (7th Ed.). Belmont, CA: Thomson
Wadsworth.
210
Willig, C. & Rogers, W.S. (2008). The SAGE handbook of qualitative research in
psychology. London: Sage Publications.
Wits University Strategic Planning Division, Facts and Figures. Retrieved December 13,
2011, from:
http://www.wits.ac.za/aboutwits/strategicplanningdivision/3153/institutional_research.h
tml
Yang, K-S, & Bond, M. H. (1990). Exploring the implicit personality theories with
indigenous or imported constructs: The Chinese case. Journal of Personality, 58, 10871095.
Yik, M.S. & Bond, M.H. (1993). Exploring the dimensions of Chinese person perception
with indigenous and imported constructs: Creating a culturally balanced scale.
International Journal of Psychology, 2, 79-95.
Zhang, L.F. &Akande, A. (2002). What relates to the big five among South Africanuniversity
students? Ife PsychlogIA: an international journal, 10, 49 – 74.
126
Table 4.3: Five and Fourteen Factor solution results for the SAPI and CPAI-2 items using varimax rotation
Factor
Scale
Subscale
1/2
Facilitating
Guidance
Encouraging Others
Soft Heartedness
Active Support
Empathy
Hostility
I help others succeed
2/3
Relationship Harmony
Harmony Maintenance
Soft Heartedness
Hostility
Egoism
Relationship Harmony
Harmony Breach
Approachability
Integrity
Integrity
Social Desirability
Social Desirability
No. of items in 14 factor
solution
35
Total: 45
6
Total: 10
7
Total: 22
1
Total: 32
1
Total: 36
1
Total: 1
8
Total: 44
25
Total: 36
2
Total 10
16
Total: 17
7
Total: 15
5
Total: 34
2
Total: 12
No. of items in 5 factor
solution
41
Total: 45
6
Total: 10
6
Total: 22
1
Total: 32
1
Total: 36
1
Total: 1
6
Total: 44
25
Total: 36
6
Total: 10
15
Total 17
8
Total: 15
7
Total: 34
3
Total: 12
127
Impression Management
3/1
Facilitating
Guidance
Soft Heartedness
Empathy
Active Support
Hostility
Egoism
Relationship Harmony
Harmony Maintenance
Approachability
4
5
Facilitating
Encouraging Others
Integrity
Integrity
Integrity
Integrity
Facilitating
Guidance
Relationship Harmony
Harmony Maintenance
Soft Heartedness
Active Support
Soft Heartedness
Egoism
Integrity
Integrity
1
Total: 1
1
Total: 45
16
Total: 32
5
Total: 22
1
Total: 36
1
Total: 10
15
Total: 44
2
Total: 15
2
Total: 10
1
Total: 34
3
Total: 34
1
Total: 45
1
Total: 44
1
Total: 22
2
Total: 10
1
Total: 34
1
Total: 6
1
Total: 45
21
Total: 32
8
Total: 22
1
Total: 36
1
Total: 10
20
Total: 44
3
Total: 15
3
Total: 10
1
Total: 34
128
6
Social Desirability
Social Desirability
Soft Heartedness
Empathy
Egoism
7
8/4
Facilitating
Encouraging Others
Relationship Harmony
Harmony Maintenance
Integrity
Integrity
Social Desirability
Social Desirability
Facilitating
Guidance
Integrity
Integrity
Soft Heartedness
Hostility
Integrity
Integrity
Social Desirability
Impression Management
1
Total: 12
3
Total: 32
2
Total: 10
1
Total: 10
1
Total: 44
1
Total: 34
1
Total: 12
2
Total: 45
2
Total: 34
1
Total: 36
4
Total: 34
2
Total: 6
Social Desirability
Self-deception
9
Relationship Harmony
Approachability
10
Relationship Harmony
Harmony Maintenance
1
Total: 15
1
Total: 44
4
Total: 34
2
Total: 12
2
Total: 7
129
Harmony Breach
11
Soft Heartedness
Egoism
Hostility
Impression Management
12
Facilitating
Guidance
13
Relationship Harmony
Harmony Maintenance
1
Total: 17
1
Total: 10
1
Total: 36
1
Total: 6
1
Total: 45
2
Total: 44